US20230205809A1 - Descriptive media content search - Google Patents
Descriptive media content search Download PDFInfo
- Publication number
- US20230205809A1 US20230205809A1 US17/993,720 US202217993720A US2023205809A1 US 20230205809 A1 US20230205809 A1 US 20230205809A1 US 202217993720 A US202217993720 A US 202217993720A US 2023205809 A1 US2023205809 A1 US 2023205809A1
- Authority
- US
- United States
- Prior art keywords
- media content
- media
- descriptive
- user
- delivery system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 claims description 105
- 238000012545 processing Methods 0.000 claims description 41
- 230000008569 process Effects 0.000 claims description 13
- 230000004044 response Effects 0.000 claims description 8
- 238000012552 review Methods 0.000 claims description 8
- 238000013500 data storage Methods 0.000 claims description 3
- 235000019640 taste Nutrition 0.000 description 38
- 230000003993 interaction Effects 0.000 description 24
- 230000002040 relaxant effect Effects 0.000 description 24
- 239000013598 vector Substances 0.000 description 22
- 238000004891 communication Methods 0.000 description 17
- 238000001514 detection method Methods 0.000 description 16
- 230000000694 effects Effects 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 10
- 230000036651 mood Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 230000001413 cellular effect Effects 0.000 description 5
- 230000005236 sound signal Effects 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 238000010845 search algorithm Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 3
- 238000003786 synthesis reaction Methods 0.000 description 3
- 241001025261 Neoraja caerulea Species 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 239000000872 buffer Substances 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 241000252794 Sphinx Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000000712 assembly Effects 0.000 description 1
- 238000000429 assembly Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000033764 rhythmic process Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/48—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2457—Query processing with adaptation to user needs
- G06F16/24578—Query processing with adaptation to user needs using ranking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/41—Indexing; Data structures therefor; Storage structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/43—Querying
- G06F16/438—Presentation of query results
- G06F16/4387—Presentation of query results by the use of playlists
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/686—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title or artist information, time, location or usage information, user ratings
Definitions
- Music search algorithms are typically based on known metadata of the music that is stored in a music library metadata database.
- the metadata includes typical information that is commonly associated with music, such as the name of the song, the name of the artist, a genre, a release date, and the like. This type of music search may be helpful for a user to identify a particular track if the user knows that information about the track. However, the music search of this type does not provide relevant search results if the user query describes something that is not included in the music metadata database.
- the present disclosure is directed to descriptive media content search.
- the descriptive media content search utilizes playlist descriptions to provide additional information to describe media content.
- playlist descriptions to provide additional information to describe media content.
- One aspect is a method of selecting and providing media content responsive to a query.
- the method includes identifying descriptive terms associated with media content items, and selecting and providing one or more media content items responsive to a query.
- the descriptive terms associated with media content items may be identified by identifying a plurality of playlists, each of the plurality of playlists associated with one or more media content items and having a playlist description; and for each of the plurality of playlists, tokenizing the playlist description into one or more tokenized terms; and associating each of the media content items in the respective playlist with the tokenized terms.
- the media content items responsive to a query may be selected and provided by receiving the query including at least one descriptor, identifying the descriptor from the query, and identifying one or more media content items, from the plurality of playlists, associated with the descriptor based on the tokenized terms.
- the user query is a voice request from a user.
- the method may include converting the voice request into text data, determining an intent and at least one slot from the text data, and identifying the descriptor from at least one of the intent and the at least one slot.
- the descriptive terms associated with media content items may be identified by, for each media content item in the plurality of playlists, weighting the media content items associated with the tokenized terms.
- the identification of the media content items may be based on the tokenized terms and the weightings of the tokenized terms.
- the method may include calculating a first term frequency, the first term frequency associated with a frequency of each of the tokenized terms appearing in each of the media content items; calculating a second term frequency, the second term frequency associated with a frequency of each of the tokenized terms appearing in all of the media content items; and calculating a relevance score based on the first term frequency and the second term frequency, the relevance score representative of relevancy of each of the tokenized terms to each of the media content items.
- the media content items are identified from the plurality of playlists based at least in part on the relevance score.
- weighting is performed using term frequency and inverse document frequency (TFIDF).
- the method may include, after weighting the media content items associated with the tokenized terms, generating an inverted index that stores the associations between the media content items with the tokenized terms.
- the playlist description includes a playlist title and a descriptive text.
- the descriptive text is obtained from at least one of user reviews, expert reviews, and statistics.
- the method may include retrieving a user profile associated with a user who provided the user query, and determining personalized media content items from the resulting media content items based on the user profile.
- the method may include retrieving a descriptive term lexicon.
- the playlist description may be tokenized into one or more tokenized terms based on the descriptive term lexicon.
- the method may include retrieving metadata of media content items of the playlists.
- the media content items may be identified from the plurality of playlists based in part on the metadata.
- the media delivery system includes at least one processing device and at least one data storage device storing data instructions executable by the at least one processing device to cause the media delivery system to identify descriptive terms associated with media content items, and select and provide one or more media content items responsive to a user query.
- the descriptive terms associated with media content items may be identified by identifying a plurality of playlists, each of the plurality of playlists associated with one or more media content items and having a playlist description; and, for each of the plurality of playlists, tokenizing the playlist description into one or more tokenized terms; and associating each of the media content items in the respective playlist with the tokenized terms.
- the media content items responsive to a user query may be selected and provided by receiving the user query including at least one descriptor, identifying the descriptor from the user query, and identifying one or more media content items, from the plurality of playlists, associated with the descriptor based on the tokenized terms.
- the data instructions may further cause the media delivery system to, for each media content item in the plurality of playlists, weight the media content items associated with the tokenized terms.
- the media content items may be identified based on the tokenized terms and the weightings of the tokenized terms.
- the data instructions may further cause the media delivery system to calculate a first term frequency, the first term frequency associated with a frequency of each of the tokenized terms appearing in each of the media content items; calculate a second term frequency, the second term frequency associated with a frequency of each of the tokenized terms appearing in all of the media content items; and calculate a relevance score based on the first term frequency and the second term frequency, the relevance score representative of relevancy of each of the tokenized terms to each of the media content items.
- the media content items may be identified from the plurality of playlists based at least in part on the relevance score.
- the data instructions may further cause the media delivery system to weight the media content items associated with the tokenized terms using term frequency and inverse document frequency (TFIDF).
- TFIDF term frequency and inverse document frequency
- the data instructions may further cause the media delivery system to, after weighting the media content items associated with the tokenized terms, generate an inverted index that stores the associations between the media content items with the tokenized terms.
- the playlist description may include a playlist title and a descriptive text.
- FIG. 1 illustrates an example system for providing media content to a user.
- FIG. 2 is a block diagram of an example media playback device of the system shown in FIG. 1 .
- FIG. 3 is a flowchart of an example method for selecting and providing media content responsive to a user query.
- FIG. 4 is a flowchart of an example method for processing an utterance of the user query.
- FIG. 5 is a flowchart of an example method for performing descriptive search for media content based on the user query.
- FIG. 6 is a flowchart of an example method for generating and updating descriptive search database.
- FIG. 7 illustrates an example method for determining candidate playlists.
- FIG. 8 illustrates an example method for identifying one or more descriptive terms from the candidate playlists.
- FIG. 9 illustrates an example data structure that associates the descriptive terms with media content items.
- FIG. 10 is a flowchart of an example method for weighting the media content items associated with the descriptive terms.
- FIG. 11 illustrates an example descriptive search database.
- FIG. 12 is a flowchart of an example method for performing a descriptive search from the descriptive search database.
- FIG. 13 illustrates an example method for analyzing the user query to identify one or more search keywords.
- FIG. 14 illustrates an example method for selecting media content items based on descriptive search utilizing the descriptive search database.
- FIG. 15 illustrates an example method for selecting personalized media content items.
- the present disclosure relates to a descriptive media content search solution that allows for searching for and identifying media content based on descriptive terms.
- a search request can include a series of descriptors.
- Such a search request with descriptors may not return satisfactory search results if used with typical entity-focused search indexes, such as search algorithms based on known metadata of the media content that is stored in a music library metadata database.
- the search request includes descriptors that are not found in the metadata, the search results do not fully reflect the descriptors in the search request.
- a descriptive term to be searched can include “relaxing” (or its variant) or “relaxing jazz” (or its variant), and this descriptive term is not typically found in the metadata of media content.
- the present disclosure provides a solution to allow a user to search for media content that better matches the user's descriptive search request. Further, the solution according to the present disclosure can provide relevant search results even when the user do not know specific information about the media content the user wants to listen to, but can describe general characteristic of the media content with descriptive terms. Examples of descriptive terms relate to moods (relaxing, happy, sad, somber, etc.), types of activities (such as running, working out, driving, dancing, etc.), or any other descriptors that are not typically found in the metadata of media content but describe some aspect of the media content.
- An exemplary embodiment of the present disclosure utilizes an extensive catalog of playlists each having a playlist description, such as a playlist title or other descriptive text, and identifies additional descriptive information for media content items to be searched.
- the embodiment of the present disclosure can involve setting up a descriptive search database, and then utilizing the descriptive search database to conduct a descriptive search responsive to a descriptive search query.
- the embodiment of the present disclosure can analyze a variety of data associated with the catalog of playlists, such as the playlist descriptions, to generate descriptive terms (e.g., tokens) for media content to be searched for.
- the descriptive search database can be generated with the descriptive terms that map to media content items in the catalog of playlists.
- FIG. 1 illustrates an example system 100 for providing media content to a user U.
- the system 100 includes a media playback device 102 , a media delivery system 104 , and a data communication network 106 .
- the media delivery system 104 includes a descriptive media content search engine 110 .
- An example user query 120 and a media output 122 are also shown.
- the media playback device 102 operates to provide media content to a user U. As described herein, the media playback device 102 operates to receive the user query 120 and provide the media output 122 to the user U according to the user query 120 . In some embodiments, the media playback device 102 operates to communicate with a system external to the media playback device 102 , such as the media delivery system 104 . The media playback device 102 can interact with the media delivery system 104 to process the user query 120 and identify media content in response to the user query 120 . In some embodiments, the media playback device 102 operates to receive the media content that is identified and provided (e.g., streamed, transmitted, etc.) by the media delivery system 104 .
- the media playback device 102 operates to play the media content and generate the media output 122 using a media output device (e.g., a speaker) therein. In other embodiments, the media playback device 102 operates to transmit the media content to another device for playback, such as an external speaker or another media playback device (e.g., a vehicle entertainment system or a home entertainment system).
- a media output device e.g., a speaker
- the media playback device 102 operates to transmit the media content to another device for playback, such as an external speaker or another media playback device (e.g., a vehicle entertainment system or a home entertainment system).
- an example of the media playback device 102 is illustrated and described in more detail herein, such as with reference to FIG. 2 .
- the media delivery system 104 operates to provide media content to one or more media playback devices, such as the media playback device 102 , via the network 106 .
- An example of the media delivery system 104 is illustrated and described in further detail herein, such as with reference to FIG. 2 .
- the network 106 is a data communication network that facilitates data communication between the media playback device 102 and the media delivery system 104 .
- the network 106 typically includes a set of computing devices and communication links between the computing devices.
- the computing devices in the network 106 use the links to enable communication among the computing devices in the network.
- the network 106 can include one or more routers, switches, mobile access points, bridges, hubs, intrusion detection devices, storage devices, standalone server devices, blade server devices, sensors, desktop computers, firewall devices, laptop computers, handheld computers, mobile telephones, vehicular computing devices, and other types of computing devices.
- the network 106 includes various types of communication links.
- the network 106 can include wired and/or wireless links, including cellular, Bluetooth®, Wi-Fi®, ultra-wideband (UWB), 802.11, ZigBee, near field communication (NFC), an ultrasonic data transmission, and other types of wireless links.
- the network 106 is implemented at various scales.
- the network 106 can be implemented as one or more vehicle area networks, local area networks (LANs), metropolitan area networks, subnets, wide area networks (WAN) (such as the Internet), or can be implemented at another scale.
- the network 106 includes multiple networks, which may be of the same type or of multiple different types.
- the media delivery system 104 includes the descriptive media content search engine 110 that operates to determine media content in response to the user query 120 .
- the user query 120 includes a search request from the user U to identify media content.
- the user query 120 can be also referred to herein as a search query, a search request, or the like.
- the user query 120 can be a text that is typed using the media playback device 102 or another computing device.
- the user query 120 can be a voice request received through a sound detection device (e.g., a microphone). As described herein, the voice request can be processed into a text query suitable for media content search.
- the user query 120 includes one or more descriptive terms (also referred to herein as descriptors) that can be used as one or more keywords to identify media content associated with the keywords.
- the user query 120 can be a voice request that is converted into a text query of “play relaxing jazz for tonight.”
- the descriptive terms in the user query 120 can include “relaxing” or its variant (e.g., “relax” or “relaxation”), “tonight” or its variant (e.g., “night”), and/or “jazz” or its variant (e.g., “jazzy”).
- the descriptive terms in the user query 120 can include “relaxing jazz” as a single phrase instead of “relaxing” and “jazz” as two separate words. Other examples are also possible.
- the descriptive media content search engine 110 operates to select media content based on one or more descriptive terms in the user query 120 .
- the descriptive media content search engine 110 can identify media content associated with at least one of the descriptive terms of the user query 120 , such as “relaxing,” “tonight,” “relaxing jazz,” or “jazz” in the illustrated example.
- An example of the descriptive media content search engine 110 is illustrated and described in more detail herein, such as with reference to FIG. 3 .
- FIG. 2 is a block diagram of an example embodiment of the media playback device 102 of the media content provision system 100 shown in FIG. 1 .
- the media playback device 102 includes a user input device 130 , a display device 132 , a wireless data communication device 134 , a media content output device 140 , a processing device 148 , and a memory device 150 .
- the media playback device 102 operates to play media content.
- the media playback device 102 is configured to play media content that is provided (e.g., streamed or transmitted) by a system external to the media playback device 102 , such as the media delivery system 104 , another system, or a peer device.
- the media playback device 102 operates to play media content stored locally on the media playback device 102 .
- the media playback device 102 operates to play media content that is stored locally as well as media content provided by other systems.
- the media playback device 102 is a handheld or portable entertainment device, smartphone, tablet, watch, wearable device, or any other type of computing device capable of playing media content.
- the media playback device 102 is a laptop computer, desktop computer, television, gaming console, set-top box, network appliance, blue-ray or DVD player, media player, stereo, or radio.
- the media playback device 102 is a system dedicated for streaming personalized media content in a vehicle environment.
- An example of such a vehicular media playback device is described in U.S. patent application Ser. No. 15/973,198, titled PERSONAL MEDIA STREAMING APPLIANCE SYSTEM, filed May 7, 2018, the disclosure of which is hereby incorporated by reference in its entirety.
- the user input device 130 operates to receive a user input 152 from a user U for controlling the media playback device 102 .
- the user input 152 can include a manual input 154 and a voice input 156 .
- the user input device 130 includes a manual input device 160 and a sound detection device 162 .
- the manual input device 160 operates to receive the manual input 154 for controlling playback of media content via the media playback device 102 .
- the manual input device 160 includes one or more buttons, keys, touch levers, switches, and/or other mechanical input devices for receiving the manual input 154 .
- the manual input device 160 includes a text entry interface, such as a mechanical keyboard, a virtual keyboard, or a handwriting input device, which is configured to receive a text input, such as a text version of the user query 120 .
- the manual input 154 is received for managing various pieces of information transmitted via the media playback device 102 and/or controlling other functions or aspects associated with the media playback device 102 .
- the sound detection device 162 operates to detect and record sounds from proximate the media playback device 102 .
- the sound detection device 162 can detect sounds including the voice input 156 .
- the sound detection device 162 includes one or more acoustic sensors configured to detect sounds proximate the media playback device 102 .
- acoustic sensors of the sound detection device 162 include one or more microphones. Various types of microphones can be used for the sound detection device 162 of the media playback device 102 .
- the voice input 156 is a user's voice (also referred to herein as an utterance) for controlling playback of media content via the media playback device 102 .
- the voice input 156 includes a voice version of the user query 120 received from the sound detection device 162 of the media playback device 102 .
- the voice input 156 is a user's voice for managing various data transmitted via the media playback device 102 and/or controlling other functions or aspects associated with the media playback device 102 .
- the sounds detected by the sound detection device 162 can be processed by the sound processing engine 180 of the media playback device 102 as described below.
- the display device 132 operates to display information to the user U. Examples of such information include media content playback information, notifications, and other information.
- the display screen 132 is configured as a touch sensitive display and includes the manual input device 160 of the user input device 130 for receiving the manual input 154 from a selector (e.g., a finger, stylus etc.) controlled by the user U. In some embodiments, therefore, the display screen 132 operates as both a display device and a user input device.
- the touch sensitive display screen 132 operates to detect inputs based on one or both of touches and near-touches.
- the display screen 132 displays a graphical user interface for interacting with the media playback device 102 .
- Other embodiments of the display screen 132 do not include a touch sensitive display screen. Some embodiments include a display device and one or more separate user interface devices. Further, some embodiments do not include a display device.
- the data communication device 134 operates to enable the media playback device 102 to communicate with one or more computing devices over one or more networks, such as the network 106
- the data communication device 134 is configured to communicate with the media delivery system 104 and receive media content from the media delivery system 104 at least partially via the network 106 .
- the data communication device 134 can be a network interface of various types which connects the media playback device 102 to the network 106 .
- Examples of the data communication device 134 include wired network interfaces and wireless network interfaces.
- Wireless network interfaces includes infrared, BLUETOOTH® wireless technology, 802.11a/b/g/n/ac, and cellular or other radio frequency interfaces in at least some possible embodiments.
- Examples of cellular network technologies include LTE, WiMAX, UMTS, CDMA2000, GSM, cellular digital packet data (CDPD), and Mobitex.
- the media content output device 140 operates to output media content.
- the media content output device 140 generates the media output 122 for the user U.
- the media content output device 140 includes one or more embedded speakers 164 which are incorporated in the media playback device 102 .
- some embodiments of the media playback device 102 include an external speaker interface 166 as an alternative output of media content.
- the external speaker interface 166 is configured to connect the media playback device 102 to another system having one or more speakers, such as headphones, a portal speaker, and a vehicle entertainment system, so that the media output 122 is generated via the speakers of the other system external to the media playback device 102 .
- Examples of the external speaker interface 166 include an audio output jack, a USB port, a Bluetooth transmitter, a display panel, and a video output jack. Other embodiments are possible as well.
- the external speaker interface 166 is configured to transmit a signal that can be used to reproduce an audio signal by a connected or paired device such as headphones or a speaker.
- the processing device 148 comprises one or more central processing units (CPU). In other embodiments, the processing device 148 additionally or alternatively includes one or more digital signal processors, field-programmable gate arrays, or other electronic circuits.
- CPU central processing units
- the processing device 148 additionally or alternatively includes one or more digital signal processors, field-programmable gate arrays, or other electronic circuits.
- the memory device 150 typically includes at least some form of computer-readable media.
- the memory device 150 can include at least one data storage device.
- Computer readable media includes any available media that can be accessed by the media playback device 102 .
- computer-readable media includes computer readable storage media and computer readable communication media.
- Computer readable storage media includes volatile and nonvolatile, removable and non-removable media implemented in any device configured to store information such as computer readable instructions, data structures, program modules, or other data.
- Computer readable storage media includes, but is not limited to, random access memory, read only memory, electrically erasable programmable read only memory, flash memory and other memory technology, compact disc read only memory, blue ray discs, digital versatile discs or other optical storage, magnetic storage devices, or any other medium that can be used to store the desired information and that can be accessed by the media playback device 102 .
- computer readable storage media is non-transitory computer readable storage media.
- Computer readable communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
- modulated data signal refers to a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal.
- computer readable communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, radio frequency, infrared, and other wireless media. Combinations of any of the above are also included within the scope of computer readable media.
- the memory device 150 operates to store data and instructions.
- the memory device 150 stores instructions for a media content cache 172 , a caching management engine 174 , a media playback engine 176 , a sound processing engine 180 , and a voice interaction engine 182 .
- the media content cache 172 stores media content items, such as media content items that have been received from the media delivery system 104 .
- the media content items stored in the media content cache 172 may be stored in an encrypted or unencrypted format.
- the media content cache 172 also stores metadata about media content items such as title, artist name, album name, length, genre, mood, era, etc.
- the media content cache 172 can further store playback information about the media content items and/or other information associated with the media content items.
- the caching management engine 174 is configured to receive and cache media content in the media content cache 172 and manage the media content stored in the media content cache 172 . In some embodiments, when media content is streamed from the media delivery system 104 , the caching management engine 174 operates to cache at least a portion of the media content into the media content cache 172 . In other embodiments, the caching management engine 174 operates to cache at least a portion of media content into the media content cache 172 while online so that the cached media content is retrieved for playback while the media playback device 102 is offline.
- the media playback engine 176 operates to play media content to the user U. As described herein, the media playback engine 176 is configured to communicate with the media delivery system 104 to receive one or more media content items (e.g., through the media stream 232 ). In other embodiments, the media playback engine 176 is configured to play media content that is locally stored in the media playback device 102 .
- the media playback engine 176 operates to retrieve one or more media content items that are either locally stored in the media playback device 102 or remotely stored in the media delivery system 104 . In some embodiments, the media playback engine 176 is configured to send a request to the media delivery system 104 for media content items and receive information about such media content items for playback.
- the sound processing engine 180 is configured to receive sound signals obtained from the sound detection device 162 and process the sound signals to identify different sources of the sounds received via the sound detection device 162 .
- the sound processing engine 180 operates to filter the user's voice input 156 (e.g., a voice request of the user query 120 ) from noises included in the detected sounds.
- Various noise cancellation technologies such as active noise control or cancelling technologies or passive noise control or cancelling technologies, can be used to filter the voice input from ambient noise.
- the sound processing engine 180 filters out omni-directional noise and preserves directional noise (e.g., an audio input difference between two microphones) in audio input.
- the sound processing engine 180 removes frequencies above or below human speaking voice frequencies.
- the sound processing engine 180 subtracts audio output of the device from the audio input to filter out the audio content being provided by the device. (e.g., to reduce the need of the user to shout over playing music). In examples, the sound processing engine 180 performs echo cancellation. By using one or more of these techniques, the sound processing engine 180 provides sound processing customized for use in a vehicle environment.
- the sound processing engine 180 operates to process the received sound signals to identify the sources of particular sounds of the sound signals, such as people's conversation in the vehicle, the vehicle engine sound, or other ambient sounds associated with the vehicle.
- a recording of sounds captured using the sound detection device 162 can be analyzed using speech recognition technology to identify words spoken by the user.
- the words may be recognized as commands from the user that alter the playback of media content and/or other functions or aspects of the media playback device 102 .
- the words and/or the recordings may also be analyzed using natural language processing and/or intent recognition technology to determine appropriate actions to take based on the spoken words.
- the sound processing engine 180 may determine various sound properties about the sounds proximate the media playback device 102 such as volume, dominant frequency or frequencies, etc. These sound properties may be used to make inferences about the environment proximate to the media playback device 102 .
- the voice interaction engine 182 operates to cooperate with the media delivery system 104 (e.g., a voice interaction server 204 thereof) to identify a command (e.g., a user intent) that is conveyed by the voice input 156 .
- the voice interaction engine 182 transmits the user's voice input 156 that is detected by the sound processing engine 180 to the media delivery system 104 so that the media delivery system 104 operates to determine a command intended by the voice input 156 .
- at least some of the determination process of the command can be performed locally by the voice interaction engine 182 .
- some embodiments of the voice interaction engine 182 can operate to cooperate with the media delivery system 104 (e.g., the voice interaction server 204 thereof) to provide a voice assistant that performs various voice-based interactions with the user, such as voice feedbacks, voice notifications, voice recommendations, and other voice-related interactions and services.
- the media delivery system 104 e.g., the voice interaction server 204 thereof
- voice assistant that performs various voice-based interactions with the user, such as voice feedbacks, voice notifications, voice recommendations, and other voice-related interactions and services.
- the media delivery system 104 includes a media content server 200 , a media content search server 202 , a voice interaction server 204 , and a user command interpretation server 206 .
- the media delivery system 104 comprises one or more computing devices and provides media content to the media playback device 102 and, in some embodiments, other media playback devices as well. In addition, the media delivery system 104 interacts with the media playback device 102 to provide the media playback device 102 with various functionalities.
- the media content server 200 , the media content search server 202 , the voice interaction server 204 , and the user command interpretation server 206 are provided by separate computing devices. In other embodiments, the media content server 200 , the media content search server 202 , the voice interaction server 204 , and the user command interpretation server 206 are provided by the same computing device(s). Further, in some embodiments, at least one of the media content server 200 , the media content search server 202 , the voice interaction server 204 , and the user command interpretation server 206 is provided by multiple computing devices. For example, the media content server 200 , the media content search server 202 , the voice interaction server 204 , and the user command interpretation server 206 may be provided by multiple redundant servers located in multiple geographic locations.
- FIG. 2 shows a single media content server 200 , a single media content search server 202 , a single voice interaction server 204 , and a single user command interpretation server 206
- some embodiments include multiple media content servers, media content search servers, voice interaction servers, and user command interpretation servers.
- each of the multiple media content servers, media content search servers, voice interaction servers, and user command interpretation servers may be identical or similar to the media content server 200 , the media content search server 202 , the voice interaction server 204 , and the user command interpretation server 206 , respectively, as described herein, and may provide similar functionality with, for example, greater capacity and redundancy and/or services from multiple geographic locations.
- some of the multiple media content servers, the media content search servers, the voice interaction servers, and/or the user command interpretation servers may perform specialized functions to provide specialized services. Various combinations thereof are possible as well.
- the media content server 200 transmits stream media to media playback devices such as the media playback device 102 .
- the media content server 200 includes a media server application 212 , a processing device 214 , a memory device 216 , and a network access device 218 .
- the processing device 214 and the memory device 216 may be similar to the processing device 148 and the memory device 150 , respectively, which have each been previously described. Therefore, the description of the processing device 214 and the memory device 216 are omitted for brevity purposes.
- the network access device 218 operates to communicate with other computing devices over one or more networks, such as the network 106 .
- Examples of the network access device include one or more wired network interfaces and wireless network interfaces.
- Examples of such wireless network interfaces of the network access device 218 include wireless wide area network (WWAN) interfaces (including cellular networks) and wireless local area network (WLANs) interfaces. In other examples, other types of wireless interfaces can be used for the network access device 218 .
- WWAN wireless wide area network
- WLANs wireless local area network
- the media server application 212 is configured to stream media content, such as music or other audio, video, or other suitable forms of media content.
- the media server application 212 includes a media stream service 222 , a media application interface 224 , and a media data store 226 .
- the media stream service 222 operates to buffer media content, such as media content items 230 A, 230 B, and 230 N (collectively 230 ), for streaming to one or more media streams 232 A, 232 B, and 232 N (collectively 232 ).
- the media application interface 224 can receive requests or other communication from media playback devices or other systems, such as the media playback device 102 , to retrieve media content items from the media content server 200 .
- the media application interface 224 receives communication from the media playback device 102 to receive media content from the media content server 200 .
- the media data store 226 stores media content items 234 , media content metadata 236 , media contexts 238 , user accounts 240 , and taste profiles 242 .
- the media data store 226 may comprise one or more databases and file systems. Other embodiments are possible as well.
- the media content items 234 may be audio, video, or any other type of media content, which may be stored in any format for storing media content.
- the media content metadata 236 provides various information associated with the media content items 234 .
- the media content metadata 236 provides various information associated with the media contexts 238 .
- the media content metadata 236 includes one or more of title, artist name, album name, length, genre, mood, era, etc.
- the media content metadata 236 operates to provide various pieces of information associated with the media content items 234 and/or the media contexts 238 .
- the media content metadata 236 includes one or more of title, artist name, album name, length, genre, mood, era, etc.
- the media content metadata 236 includes acoustic metadata, cultural metadata, and explicit metadata.
- the acoustic metadata may be derived from analysis of the track and refers to a numerical or mathematical representation of the sound of a track.
- Acoustic metadata may include temporal information such as tempo, rhythm, beats, downbeats, tatums, patterns, sections, or other structures.
- Acoustic metadata may also include spectral information such as melody, pitch, harmony, timbre, chroma, loudness, vocalness, or other possible features.
- Acoustic metadata may take the form of one or more vectors, matrices, lists, tables, and other data structures.
- Acoustic metadata may be derived from analysis of the music signal.
- One form of acoustic metadata commonly termed an acoustic fingerprint, may uniquely identify a specific track. Other forms of acoustic metadata may be formed by compressing the content of a track while retaining some or all of its musical characteristics.
- the cultural metadata refers to text-based information describing listeners' reactions to a track or song, such as styles, genres, moods, themes, similar artists and/or songs, rankings, etc.
- Cultural metadata may be derived from expert opinion such as music reviews or classification of music into genres.
- Cultural metadata may be derived from listeners through websites, chatrooms, blogs, surveys, and the like.
- Cultural metadata may include sales data, shared collections, lists of favorite songs, and any text information that may be used to describe, rank, or interpret music.
- Cultural metadata may also be generated by a community of listeners and automatically retrieved from Internet sites, chat rooms, blogs, and the like.
- Cultural metadata may take the form of one or more vectors, matrices, lists, tables, and other data structures.
- a form of cultural metadata particularly useful for comparing music is a description vector.
- a description vector is a multi-dimensional vector associated with a track, album, or artist. Each term of the description vector indicates the probability that a corresponding word or phrase would be used to describe the associated track, album or artist.
- the explicit metadata refers to factual or explicit information relating to music.
- Explicit metadata may include album and song titles, artist and composer names, other credits, album cover art, publisher name and product number, and other information.
- Explicit metadata is generally not derived from the music itself or from the reactions or opinions of listeners.
- At least some of the metadata 236 such as explicit metadata (names, credits, product numbers, etc.) and cultural metadata (styles, genres, moods, themes, similar artists and/or songs, rankings, etc.), for a large library of songs or tracks can be evaluated and provided by one or more third party service providers.
- Acoustic and cultural metadata may take the form of parameters, lists, matrices, vectors, and other data structures. Acoustic and cultural metadata may be stored as XML files, for example, or any other appropriate file type.
- Explicit metadata may include numerical, text, pictorial, and other information. Explicit metadata may also be stored in an XML, or other file. All or portions of the metadata may be stored in separate files associated with specific tracks. All or portions of the metadata, such as acoustic fingerprints and/or description vectors, may be stored in a searchable data structure, such as a k-D tree or other database format.
- each of the media contexts 238 is used to identify one or more media content items 234 .
- the media contexts 238 are configured to group one or more media content items 234 and provide a particular context to the group of media content items 234 .
- Some examples of the media contexts 238 include albums, artists, playlists, and individual media content items.
- the media context 238 can represent that the media content items 234 identified by the media context 238 are associated with that album.
- the media contexts 238 can include playlists 239 .
- the playlists 239 are used to identify one or more of the media content items 234 .
- the playlists 239 identify a group of the media content items 234 in a particular order.
- the playlists 239 merely identify a group of the media content items 234 without specifying a particular order.
- Some, but not necessarily all, of the media content items 234 included in a particular one of the playlists 239 are associated with a common characteristic such as a common genre, mood, or era.
- a user can listen to media content items in a playlist 239 by selecting the playlist 239 via a media playback device, such as the media playback device 102 .
- the media playback device then operates to communicate with the media delivery system 104 so that the media delivery system 104 retrieves the media content items identified by the playlist 239 and transmits data for the media content items to the media playback device for playback.
- the playlist 239 includes one or more playlist descriptions, such as a playlist description 570 as illustrated in FIG. 8 .
- the playlist descriptions include information associated with the playlist 239 .
- the playlist descriptions can include a playlist title, such as a playlist title 574 as illustrated in FIG. 8 .
- the playlist title is a title of the playlist.
- the playlist title can be provided by a user using the media playback device 102 .
- the playlist title can be provided by a media content provider (or a media-streaming service provider).
- the playlist title can be automatically generated.
- playlist descriptions include a descriptive text, such as a descriptive text 576 as illustrated in FIG. 8 .
- the descriptive text can be provided by the user and/or the media content provider, which is to represent the corresponding playlist 239 .
- the descriptive text of the playlist description can be obtained from one or more other sources.
- Such other sources can include expert opinion (e.g., music reviews or classification of music into genres), user opinion (e.g., reviews through websites, chatrooms, blogs, surveys, and the like), statistics (e.g., sales data), shared collections, lists of favorite playlists, and any text information that may be used to describe, rank, or interpret the playlist or music associated with the playlist.
- the playlist descriptions can also be generated by a community of listeners and automatically retrieved from Internet sites, chat rooms, blogs, and the like.
- the playlist descriptions can take the form of one or more vectors, matrices, lists, tables, and other data structures.
- a form of cultural metadata particularly useful for comparing music is a description vector.
- a description vector is a multi-dimensional vector associated with a track, album, or artist. Each term of the description vector indicates the probability that a corresponding word or phrase would be used to describe the associated track, album or artist. Each term of the description vector indicates the probability that a corresponding word or phrase would be used to describe the associated track, album or artist.
- the playlist 239 includes a list of media content item identifications (IDs), such as media content item identifications (IDs) 578 as illustrated in FIG. 8 .
- the list of media content item identifications includes one or more media content item identifications that refer to respective media content items 234 .
- Each media content item is identified by a media content item ID and includes various pieces of information, such as a media content item title, artist identification (e.g., individual artist name or group name, or multiple artist names or group names), and media content item data.
- the media content item title and the artist ID are part of the media content metadata 236 , which can further include other attributes of the media content item, such as album name, length, genre, mood, era, etc. as described herein.
- At least some of the playlists 239 may include user-created playlists.
- a user of a media streaming service provided using the media delivery system 104 can create a playlist 239 and edit the playlist 239 by adding, removing, and rearranging media content items in the playlist 239 .
- a playlist 239 can be created and/or edited by a group of users together to make it a collaborative playlist.
- user-created playlists can be available to a particular user only, a group of users, or to the public based on a user-definable privacy setting.
- the media delivery system 104 when a playlist is created by a user or a group of users, the media delivery system 104 operates to generate a list of media content items recommended for the particular user or the particular group of users. In some embodiments, such recommended media content items can be selected based at least on the taste profiles 242 as described herein. Other information or factors can be used to determine the recommended media content items. Examples of determining recommended media content items are described in U.S. patent application Ser. No. 15/858,377, titled MEDIA CONTENT ITEM RECOMMENDATION SYSTEM, filed Dec. 29, 2017, the disclosure of which is hereby incorporated by reference in its entirety.
- the playlists 239 are created by a media streaming service provider.
- provider-created playlists can be automatically created by the media delivery system 104 .
- a provider-created playlist can be customized to a particular user or a particular group of users.
- a playlist for a particular user can be automatically created by the media delivery system 104 based on the user's listening history (e.g., the user's taste profile) and/or listening history of other users with similar tastes.
- a provider-created playlist can be configured to be available for the public in general. Provider-created playlists can also be sharable with other users.
- the user accounts 240 are used to identify users of a media streaming service provided by the media delivery system 104 .
- a user account 240 allows a user to authenticate to the media delivery system 104 and enable the user to access resources (e.g., media content items, playlists, etc.) provided by the media delivery system 104 .
- the user can use different devices to log into the user account and access data associated with the user account in the media delivery system 104 .
- User authentication information such as a username, an email account information, a password, and other credentials, can be used for the user to log into his or her user account.
- user data is handled according to robust privacy and data protection policies and technologies. For instance, whenever personally identifiable information and any other information associated with users is collected and stored, such information is managed and secured using security measures appropriate for the sensitivity of the data. Further, users can be provided with appropriate notice and control over how any such information is collected, shared, and used.
- the taste profiles 242 contain records indicating media content tastes of users.
- a taste profile can be associated with a user and used to maintain an in-depth understanding of the music activity and preference of that user, enabling personalized recommendations, taste profiling and a wide range of social music applications.
- Libraries and wrappers can be accessed to create taste profiles from a media library of the user, social website activity and other specialized databases to obtain music preferences.
- each taste profile 242 is a representation of musical activities, such as user preferences and historical information about the users' consumption of media content, and can include a wide range of information such as artist plays, song plays, skips, dates of listen by the user, songs per day, playlists, play counts, start/stop/skip data for portions of a song or album, contents of collections, user rankings, preferences, or other mentions received via a client device, or other media plays, such as web sites visited, book titles, movies watched, playing activity during a movie or other presentations, ratings, or terms corresponding to the media, such as “comedy,” etc.
- the taste profiles 242 can include other information.
- the taste profiles 242 can include libraries and/or playlists of media content items associated with the user.
- the taste profiles 242 can also include information about the user's relationships with other users (e.g., associations between users that are stored by the media delivery system 104 or on a separate social media site).
- the taste profiles 242 can be used for a number of purposes.
- One use of taste profiles is for creating personalized playlists (e.g., personal playlisting).
- An API call associated with personal playlisting can be used to return a playlist customized to a particular user.
- the media content items listed in the created playlist are constrained to the media content items in a taste profile associated with the particular user.
- Another example use case is for event recommendation.
- a taste profile can be created, for example, for a festival that contains all the artists in the festival. Music recommendations can be constrained to artists in the taste profile.
- Yet another use case is for personalized recommendation, where the contents of a taste profile are used to represent an individual's taste.
- This API call uses a taste profile as a seed for obtaining recommendations or playlists of similar artists.
- a bulk resolution API call is used to resolve taste profile items to pre-stored identifiers associated with a service, such as a service that provides metadata about items associated with the taste profile (e.g., song tempo for a large catalog of items).
- a service such as a service that provides metadata about items associated with the taste profile (e.g., song tempo for a large catalog of items).
- user-to-user recommendation is used to discover users with similar tastes by comparing the similarity of taste profile item(s) associated with users.
- a taste profile 242 can represent a single user or multiple users. Conversely, a single user or entity can have multiple taste profiles 242 . For example, one taste profile can be generated in connection with a user's media content play activity, whereas another separate taste profile can be generated for the same user based on the user's selection of media content items and/or artists for a playlist.
- the media content search server 202 operates to perform media content search in response to a media content search request, such as the user query 120 ( FIG. 1 ).
- the media content search server 202 includes a media content search application 250 , a processing device 252 , a memory device 254 , and a network access device 256 .
- the processing device 252 , the memory device 254 , and the network access device 256 may be similar to the processing device 214 , the memory device 216 , and the network access device 218 , respectively, which have each been previously described.
- the media content search application 250 operates to interact with the media playback device 102 and provide selection of one or more media content items based on the user query 120 .
- the media content search application 250 can include the descriptive media content search engine 110 ( FIG. 1 ).
- the media content search application 250 can interact with other servers, such as the media content server 200 , the voice interaction server 204 , and the user command interpretation server 206 , to perform media content search.
- the voice interaction server 204 operates to provide various voice-related functionalities to the media playback device 102 .
- the voice interaction server 204 includes a voice recognition application 270 , a speech synthesis application 272 , a processing device 274 , a memory device 276 , and a network access device 278 .
- the processing device 274 , the memory device 276 , and the network access device 278 may be similar to the processing device 214 , the memory device 216 , and the network access device 218 , respectively, which have each been previously described.
- the voice recognition application 270 and the speech synthesis application 272 operate to interact with the media playback device 102 and enable the media playback device 102 to perform various voice-related functions, such as voice media content search, voice feedback, voice notifications, etc.
- the voice recognition application 270 is configured to perform speech-to-text (STT) conversion, such as receiving a recording of voice command (e.g., an utterance) and converting the utterance to a text format.
- STT speech-to-text
- the speech synthesis application 272 is configured to perform text-to-speech (TTS) conversion, so that a language text is converted into speech. Then, the voice interaction server 204 can transmit an audio data or file for the speech to the media playback device 102 so that the media playback device 102 generates a voice assistance to the user using the transmitted audio data or file.
- TTS text-to-speech
- the user command interpretation server 206 operates to analyze a text version of a user command (e.g., a text version of the utterance) to determine appropriate actions to take according to the user command.
- a text version of a user command e.g., a text version of the utterance
- the user command interpretation server 206 includes a natural language understanding (NLU) application 280 , a processing device 282 , a memory device 284 , and a network access device 286 .
- NLU natural language understanding
- the processing device 282 , the memory device 284 , and the network access device 286 may be similar to the processing device 214 , the memory device 216 , and the network access device 218 , respectively, which have each been previously described.
- the NLU application 280 operates to analyze the text format of the utterance to determine functions to perform based on the utterance.
- the NLU application 280 can use a natural language understanding algorithm that involves modeling human reading comprehension, such as parsing and translating an input according to natural language principles.
- FIG. 3 is a flowchart of an example method 300 for selecting and providing media content responsive to a user query 120 . It is described herein that the method 300 is at least partially performed primarily by the media delivery system 104 (e.g., the descriptive media content search engine 110 thereof) that interacts with the media playback device 102 . In other embodiments, the method 300 can be at least partially performed by the media playback device 102 that interacts with the media delivery system 104 . In yet other embodiments, the method 300 can be at least partially performed by the media playback device 102 and the media delivery system 104 .
- the media delivery system 104 e.g., the descriptive media content search engine 110 thereof
- the media delivery system 104 operates to receive a media content search query, such as the user query 120 ( FIG. 1 ).
- the user query 120 can be received using the media playback device 102 .
- the user query 120 is a text typed by the user using the media playback device 102 .
- the user can type in the text (a text form of “play relaxing jazz for tonight”) via the manual input device 160 of the media playback device 102 .
- the text of the user query 120 can be analyzed to identify descriptive terms (e.g., descriptors) that are to be used for media content search as described herein.
- the user query 120 is an utterance that is received and recorded using the media playback device 102 .
- the user can speak the voice request (an audio form of “play relaxing jazz for tonight”) via the sound detection device 162 of the media playback device 102 .
- the utterance of the user query 120 received via the media playback device 102 can be processed into a text version and further analyzed to identify descriptive terms (e.g., descriptors) that can be used for media content search as described herein.
- An example method for processing the utterance of the user query 120 is illustrated and described in more detail herein, for example with reference to FIG. 4 .
- the media delivery system 104 operates to perform descriptive media content search based on the search query (e.g., the user query 120 ).
- the descriptive media content search can identify media content (e.g., one or more media content items 234 ) that is most relevant to the descriptive terms (e.g., descriptors) in the user query 120 .
- other criteria can be used to determine media content responsive to the user query 120 .
- An example method for performing such descriptive media content search is illustrated and described in more detail herein, for example with reference to FIG. 5 .
- the media delivery system 104 operates to provide the identified media content to the media playback device 102 .
- the media delivery system 104 operates to transmit media stream 232 for the identified media content to the media playback device 102 via the network 106 .
- FIG. 4 is a flowchart of an example method 400 for processing an utterance of the user query 120 .
- the method 400 can also be performed to determine a descriptive media content search service as a fulfillment strategy based on the utterance of the user query 120 .
- the method 400 can be used to at least partially perform the operation 302 of the method 300 in FIG. 3 .
- the method 400 can be performed by the media delivery system 104 .
- the method 400 can be used by the natural language understanding (NLU) system (e.g., the NLU application 280 ) for performing a fulfillment strategy based on a received utterance.
- the media delivery system 104 can include the user command interpretation server 206 that operates the NLU application 280 , and the media playback device 102 can receive an utterance of the user query 120 and provide the utterance to the media delivery system 104 for processing at the user command interpretation server 206 .
- the NLU application 280 is locally saved and performed in the media playback device 102 .
- the method 400 can begin at operation 410 in which the media delivery system 104 receives the utterance of the user query 120 .
- the utterance of the user query 120 can be stored and received as utterance data 412 by the media delivery system 104 .
- the utterance data 412 is data describing the utterance of the user query 120 .
- the utterance data 412 is an audio recording that contains the utterance being spoken.
- the utterance data 412 is received as an entire audio data file.
- the media playback device 102 buffers the utterance data 412 as it is obtained from the sound detection device 162 .
- the buffered utterance data 412 is then sent to the media delivery system 104 for processing.
- the media playback device 102 streams the utterance data 412 to the media delivery system 104 in real-time as the utterance data 412 is received from the sound detection device 162 of the media playback device 102 .
- the utterance data 412 is stored (e.g., by the media delivery system 104 ) in a data store after it is received. After the utterance data 412 is received, the method 400 moves to operation 420 .
- Operation 420 includes performing automated speech recognition on the utterance data 412 to obtain text data 422 .
- performing automated speech recognition includes providing the utterance data 412 as input to an automated speech recognition system (e.g., the voice recognition application 270 of the voice interaction server 204 ) and receiving the text data 422 as output from the automated speech recognition system.
- Automated speech recognition can be performed using any of a variety of techniques (e.g., using hidden Markov models or neural networks). Examples of automated speech recognition systems include CMU SPHINX, maintained by CARNEGIE MELLON UNIVERSITY, and DEEPSPEECH, maintained by the MOZILLA FOUNDATION.
- Operation 430 includes determining a slot 434 and an intent 432 from the text data 422 .
- the slot 434 is a key-value pair that describes a portion of the text data 422 having a specific meaning.
- the intent 432 describes a general intent of the text data 422 .
- the slot 434 can include at least one key-value pair, such as ⁇ descriptor:relaxing ⁇ and ⁇ descriptor:tonight ⁇ .
- the intent 432 is “play” and the slot 434 is a key-value pair of ⁇ song: Thriller ⁇ .
- the text data 422 can include a single slot 434 and a single intent 432 .
- the operation 430 can return a plurality of slots 434 and/or a plurality of intents 432 from the text data 422 .
- the text data 422 provides an intent 432 but no slot 434 .
- the operation 430 can return the intent 432 being “play”, but will not result in any slot 434 (e.g., the text data 422 does not include a description of what to play).
- the text data 433 can include one or more slots 434 but no intent.
- the operation 430 can result in two slots 434 (e.g., ⁇ Song: All Along the Watchtower ⁇ and ⁇ Artist: Jimi Hendrix ⁇ ) but no intent 432 (e.g., the text data 422 does not include a description of what do to with the song and artist, such as search, play, or save).
- the text data 422 can include one or more descriptors.
- the descriptors can include values of one or more of the slots 434 identified from the text data 422 .
- the values of the slots 434 such as “relaxing,” “relaxing jazz,” and/or “tonight,” can be included as descriptors of the text data 422 .
- the descriptors of the text data 422 can include the value of the intent 432 (e.g., “play” in the above example).
- other terms identified in the text data 422 can be included in the descriptors of the text data 422 .
- the descriptors of the text data 422 can be identified in other methods.
- the operation 430 can be performed by a natural language understanding model that is trained to identify the slot 434 and intent 432 for the text data 422 provided as input.
- the natural language understanding model can be implemented in a variety of ways, including using a state vector machine or a conditional random fields model, among others. With the intent 432 and the slots 434 determined, the method 400 can move to operation 440 .
- Operation 440 includes determining a fulfillment strategy 442 using the slot 434 and the intent 432 .
- the fulfillment strategy 442 is a course of action to be performed to execute a command or service associated with the utterance, such as the intent 432 thereof.
- the fulfillment strategy 442 can include a descriptive search service in response to the user query 120 .
- the fulfillment strategy 442 is a play fulfillment strategy and involves execution of a play command.
- the media delivery system 104 includes a fulfillment manager that operates to select a fulfillment strategy 442 among a plurality of fulfillment strategies, which satisfies the received utterance.
- the plurality of fulfillment strategies can include a play fulfillment strategy, a basic search strategy (e.g., using entity-focused search), a descriptive search strategy, a recommendation strategy, and an unspecified search strategy.
- the fulfillment manager follows a decision tree based the intent 432 and the slot 434 .
- each fulfillment strategy 442 defines requirements (e.g., a play fulfillment strategy may require a play intent) and the fulfillment manager selects a fulfillment strategy among the plurality of fulfillment strategies based on whether the associated requirement is met.
- the fulfillment strategy 442 is then performed by the media delivery system 104 .
- the media delivery system 104 performs descriptive media content search based on the text data 422 (e.g., at least one of the descriptors identified from the text data 422 ).
- FIG. 5 is a flowchart of an example method 500 for performing descriptive search for media content based on the user query 120 .
- the method 500 can be used to at least partially perform the operation 304 of the method 300 as illustrated in FIG. 3 .
- the media delivery system 104 can operate to identify descriptive terms associated with candidate media content items.
- the candidate media content items can include a subset of media content items from media content item database and are subject to the search in response to the user query 120 .
- the candidate media content items are obtained from candidate playlists 558 ( FIG. 7 ).
- the operation 502 is performed to set up a descriptive search database that is to be utilized to conduct a descriptive search. In some embodiments, the operation 502 can be performed to build and store the descriptive search database prior to receiving the user query 120 . In other embodiments, the operation 502 is performed upon receiving the user query 120 .
- An example of the operation 502 is illustrated and described in more detail herein, for example with reference to FIG. 6 .
- the media delivery system 104 can operate to select one or more media content items from the candidate media content items.
- the operation 504 is performed to utilize the descriptive search database and conduct a descriptive search based on one or more descriptors of the user query 120 .
- An example of the operation 504 is illustrated and described in more detail herein, for example with reference to FIG. 12 .
- FIG. 6 is a flowchart of an example method 530 for generating or updating descriptive search database.
- the method 530 is used to at least partially perform the operation 502 as illustrated in FIG. 5 .
- the descriptive search database can be generated by identifying descriptive terms associated with media content items.
- the descriptive search solution utilizes a corpus of playlists, each of which contains a playlist description that describe the respective playlist.
- the playlist description of a playlist can be used to identify additional descriptive information about media content items in the playlist, and use that information to provide relevant search results.
- the media delivery system 104 operates to retrieve candidate playlists 558 ( FIG. 7 ).
- the candidate playlists 558 includes a subset of playlists selected from playlist database and subject to the search as described herein.
- An example of the operation 532 is illustrated and described in more detail herein, for example with reference to FIG. 7 .
- the media delivery system 104 operates to retrieve a descriptive term lexicon 560 as illustrated in FIG. 8 .
- the descriptive term lexicon 560 includes a list of words or phrases that is prepopulated and used as a vocabulary to identify descriptive terms from playlist descriptions 570 in the candidate playlists 558 (including 558 A, 558 B, and 558 C), as illustrated in FIG. 8 .
- the descriptive term lexicon 560 can be used to clarify unclear terms in the playlist descriptions of the candidate playlists 558 and identify them as descriptive terms relevant to the descriptive search.
- the descriptive term lexicon 560 can be manually created and updated.
- the descriptive term lexicon 560 can be at least partially automatically created and updated using, for example, machine learning technology.
- An example of the operation 534 is illustrated and described in more detail herein, for example with reference to FIG. 8 .
- the operation 534 can be optional and, in other embodiments, can be omitted from the method 530 .
- the media delivery system 104 operates to analyze a playlist description 570 of each candidate playlist 558 and identify one or more descriptive terms 572 (including 572 A, 572 B, and 572 C) from the playlist description 570 .
- the playlist description 570 is analyzed using the descriptive term lexicon 560 .
- the playlist description 570 is tokenized into tokenized terms (e.g., tokens), and at least one of the tokenized terms can be determined as the descriptive terms 572 .
- tokenized terms e.g., tokens
- the media delivery system 104 operates to associate the descriptive terms 572 identified from each candidate playlist 558 with each of the media content items 234 in the candidate playlist 558 .
- the media content items 234 associated with the candidate playlist 558 can be identified by media content item identifications (IDs) 578 .
- the descriptive terms 572 associated with the media content items are used as additional information descriptive of the media content items.
- An example association of the descriptive terms 572 with the media content items is illustrated and described in more detail herein, for example with reference to FIG. 9 .
- the media delivery system 104 operates to weight the media content items 234 associated with the descriptive terms 572 .
- each media content item 234 associated with a given descriptive term 572 can be weighted to provide an indication of how strongly the media content item 234 is associated with the descriptive term 572 .
- the weighting can also be based at least in part on how frequency the media content item 234 appears in the candidate playlists that contain the descriptive term 572 in the playlist description 570 .
- a term frequency and inverse document frequency (TFIDF) weighting is used to compute a weighting score for each of the media content items 234 from the candidate playlists 558 .
- An example of the operation 540 is illustrated and described in more detail herein, for example with reference to FIG. 10 .
- the operation 540 can be optional and, in other embodiments, can be omitted from the method 530 .
- the media delivery system 104 operates to generate and/or update a descriptive search database 610 ( FIG. 11 ).
- the descriptive terms 572 can be stored in the descriptive search database 610 .
- an inverted index is created to map the descriptive terms 572 to the media content items 234 in the candidate playlists 558 .
- the descriptive search database 610 stores each of the descriptive terms 572 and includes a list of media content items associated with that descriptive term 572 .
- An example data structure of the descriptive search database 610 is illustrated and described in more detail herein, for example with reference to FIG. 11 .
- FIG. 7 illustrates an example method 548 for determining candidate playlists 558 .
- the method 548 is used to at least partially perform the operation 532 as described in FIG. 6 .
- the method 548 can be performed to identify a subset of playlists from a playlist database 552 .
- the playlist database 552 includes a plurality of playlists 554 , which can include the playlists 239 as described in FIG. 2 .
- the subset of playlists can be referred to herein as candidate playlists 558 .
- a candidate playlist database 556 is created to store the candidate playlists 558 .
- the method 548 is designed to filter out some of the playlists 554 in the playlist database 552 that have less coherent or suitable to be used for descriptive search.
- some of the playlists 554 in the playlist database 552 can have no or little playlist description, or playlist descriptions with no meaningful information that characterize the playlists. These playlists 554 in the playlist database 552 are not useful for performing the method 530 as described herein.
- the method 548 is performed to select candidate playlists 558 that are meaningful, suitable playlists that can lead to relevant search results in response to the user query 120 .
- the method 548 can be performed at least by a candidate playlist selection engine 550 that can be included in the media delivery system 104 .
- the candidate playlist selection engine 550 can operate to retrieve the playlists 554 stored in the playlist database 552 . In other embodiments, only some of the playlists 554 in the playlist database 552 are retrieved.
- the candidate playlist selection engine 550 can use one or more various algorithms to select candidate playlists 558 .
- a heuristic technique can be used to select candidate playlists 558 from the playlists 554 .
- the candidate playlists 558 can be selected based on a predetermined rule.
- the candidate playlists 558 in the candidate playlist database 556 can be updated as the playlists 554 in the playlist database 552 change.
- the playlists 554 can change in the playlist database 552 as users create new playlists, edit or remove existing playlists, and add, remove, and modify media content items in existing playlists.
- the candidate playlists 558 can be updated periodically, in real-time, and/or at preset times.
- FIG. 8 illustrates an example method 562 for identifying one or more descriptive terms 572 from the candidate playlists 558 .
- the method 562 is used to at least partially perform the operations 534 and 536 as described in FIG. 6 .
- a candidate playlist 558 has a playlist description 570 .
- the playlist description 570 includes a playlist title 574 and a descriptive text 576 .
- the playlist title 574 is a name of the playlist, which can be created either by a user or by a media streaming service provider (either manually or automatically).
- the descriptive text 576 can include words, phrases, or sentences that characterize the candidate playlist 558 .
- the descriptive text 576 can be obtained from various sources. In some examples, a user can provide the descriptive text 576 .
- the descriptive text 576 can be obtained from external sources, such as user or expert opinions or feedback (e.g., reviews or genre labels through websites, chatrooms, blogs, surveys, and the like), statistics (e.g., sales data), or any text information that may be used to describe, rank, or interpret the playlist or music associated with the playlist.
- the descriptive text 576 can also be generated by a community of listeners and automatically retrieved from Internet sites, chat rooms, blogs, and the like.
- the playlist description 570 can be analyzed to identify one or more descriptive terms 572 .
- the playlist description 570 is tokenized into tokenized terms (e.g., tokens), at least some of which are determined as descriptive terms 572 .
- the playlist description 570 is analyzed such that one or more keywords in the playlist description 570 are identified and stored as tokens.
- the descriptive term lexicon 560 is used in identifying keywords from the playlist description 570 and determining the keywords as the descriptive terms 572 .
- a first candidate playlist 558 A contains a playlist description 570 including a playlist title 574 (“Relaxing Night”) and a descriptive text 576 (“Great for relaxation afterwork”).
- the media delivery system 104 can process (e.g., tokenize) the playlist description 570 to identify keywords, such as “relaxing,” “night,” “great,” “relaxation,” and “afterwork” and generate tokens for such keywords and/or their variants. At least one of the tokens can be included in the descriptive terms 572 A for the first candidate playlist 558 A.
- Second and third candidate playlists 558 B and 558 C are similarly analyzed to generate the descriptive terms 572 B and 572 C, respectively.
- the playlist description 570 can be tokenized in various ways. In some embodiments, the playlist description 570 can be tokenized by a single word. In other embodiments, the playlist description 570 can be tokenized by a combination of adjacent words if such adjacent words are determined to be related. Other methods are also possible.
- FIG. 9 illustrates an example data structure 580 that associates the descriptive terms 572 with media content items 234 .
- the data structure 580 is generated from the operation 538 as illustrated in FIG. 6 .
- the data structure 580 is configured as a data table having an item title column 582 , an item ID column 584 , a playlist ID column 586 , a descriptive term column 588 , and one or more metadata columns 590 .
- the item title column 582 shows the titles of the media content items 234 .
- the item ID column 584 shows the ID (e.g., a uniform resource identifier (URI)) of the media content items 234 .
- the playlist ID column 586 shows the IDs of the playlists that contain respective media content items 234 .
- the descriptive term column 588 shows the descriptive terms 572 associated with respective media content items 234 . As described herein, the descriptive terms 572 for a given media content item 234 can be identified from the playlists (e.g., the candidate playlists 558 ) containing that media content item 234 .
- each of the media content items 234 from the candidate playlists 558 are associated with the descriptive terms 572 identified from the candidate playlists 558 .
- a media content item 234 titled “ABC” is identified as item ID “123456abw” and can be associated with the descriptive terms, “relax” or its variants (e.g., “relaxing” or “relaxation”), “great,” “night” or its variants (e.g., “tonight”), “afterwork,” “kid,” “favorite,” and “good,” which are identified from the candidate playlists 558 A, 558 B, and 558 C.
- This media content item 234 is associated with the descriptive term “relax” or its variants twice from the candidate playlists 558 A, 558 B, and 558 C, and also associated with the descriptive term “night” or its variants twice from the candidate playlists 558 A, 558 B, and 558 C.
- the association of the descriptive terms 572 identified from the candidate playlists 558 with the media content items 234 in the candidate playlists 558 provides additional descriptive information to the media content items 234 .
- the idea of this association is that the descriptive terms of a playlist is an indication of the characteristics of the media content items contained in that playlist. Therefore, a playlist description of a playlist can suggest the characteristics of the media content items in the playlist.
- the data structure 580 also includes at least part of the information from metadata (e.g., the metadata 236 ( FIG. 2 )) of the media content items 234 .
- metadata e.g., the metadata 236 ( FIG. 2 )
- Such metadata information is shown in one or more of the metadata columns 590 .
- the metadata columns 590 includes a column for genre of the media content items 234 .
- FIG. 10 is a flowchart of an example method 600 for weighting the media content items 234 associated with the descriptive terms 572 .
- the method 600 is used to at least partially perform the operation 540 as described in FIG. 6 .
- the method 600 can be performed based on the data that map the media content items 234 to the descriptive terms 572 , such as shown in the data structure 580 in FIG. 9 .
- the method 600 can be used to compute how relevant a given media content item 234 is to the user query 120 .
- the method 600 can provide a solution to ranking the media content items 234 in view of the user query 120 .
- the method 600 can determine how relevant or important a given descriptive term is to a given media content item 234 among a plurality of media content items 234 .
- the method 600 can begin at operation 602 in which the media delivery system 104 operates to calculate a first term frequency 640 .
- the first term frequency 640 is a term frequency in each of the media content items 234 .
- the first term frequency 640 is a measure of how frequently a particular term (e.g., a descriptive term) appears in a document (e.g., a media content item).
- the first term frequency (TF) 640 can be computed as:
- TF ⁇ ( descriptive ⁇ term , media ⁇ content ⁇ item ) Num ⁇ ber ⁇ of ⁇ tim ⁇ es ⁇ the ⁇ descrptive ⁇ term appears ⁇ in ⁇ the ⁇ media ⁇ content ⁇ item Total ⁇ number ⁇ of ⁇ desciptive ⁇ terms ⁇ in the ⁇ media ⁇ conten ⁇ item
- the media delivery system 104 operates to calculate a second term frequency 642 .
- the second term frequency 642 is a measure of how frequently a particular term (e.g., a descriptive term) appears in all documents (e.g., all candidate media content items).
- the second term frequency (IDF) 642 can be computed as:
- IDF ⁇ ( descriptive ⁇ term ) ln ⁇ ( Total ⁇ number ⁇ of ⁇ media ⁇ content ⁇ items Number ⁇ of ⁇ media ⁇ content ⁇ items ⁇ containing ⁇ the ⁇ term )
- IDF 0 if the term does not appear in any media content item.
- the media delivery system 104 operates to calculate a relevance score 644 based on the first term frequency 640 and the second term frequency 642 .
- the relevance score 644 is a measure of how relevant a particular term (e.g., a descriptive term) is to a document (e.g., a media content item).
- the relevance score 644 can be computed as:
- the method 600 implements a term frequency and inverse document frequency (TFIDF) weighting. In other embodiments, other weighting algorithms can be used for the method 600 .
- TFIDF term frequency and inverse document frequency
- FIG. 11 illustrates an example of the descriptive search database 610 .
- the descriptive search database 610 is generated and/or updated at least partially by the operation 542 as described in FIG. 6 .
- the descriptive search database 610 includes an inverted index structure 620 that maps the descriptive terms 572 to the media content items 234 that are contained in the candidate playlists 558 .
- Other indexing methods are also possible in the descriptive search database 610 .
- the inverted index structure 620 is configured as a data table having a descriptive term column 622 , an item title column 624 , an item ID column 626 , a relevance score column 628 , and one or more metadata columns 630 .
- the descriptive term column 622 includes the descriptive terms 572 identified from the candidate playlists 558 .
- the item title column 624 shows the titles of the media content items 234 .
- the item ID column 626 shows the ID (e.g., a uniform resource identifier (URI)) of the media content items 234 .
- the relevance score column 628 includes relevance scores calculated for each descriptive term 572 with respect to each media content item 234 . In some embodiments, the relevance scores can be obtained by the method 600 as described in FIG. 10 .
- the metadata columns 630 includes at least part of the information from metadata (e.g., the metadata 236 ( FIG. 2 )) of the media content items 234 .
- the metadata of the media content items 234 can be additionally used to improve the search results.
- FIG. 12 is a flowchart of an example method 700 for performing a descriptive search from the descriptive search database 610 .
- the method 700 is used to at least partially perform the operation 504 as illustrated in FIG. 5 .
- the method 700 is used to select one or more media content items from the candidate media content items (e.g., the media content items from the candidate playlists 558 ) by utilizing the descriptive search database and conducting a descriptive search based on one or more descriptors of the user query 120 .
- the method 700 can begin at operation 702 in which the media delivery system 104 operates to process the user query 120 to identify one or more search keywords 734 ( FIG. 13 ).
- the user query 120 is analyzed to identify one or more descriptors 736 , and one or more of the descriptors 736 from the user query 120 can be used for one or more search keywords 734 .
- An example of the operation 702 is illustrated and described in more detail herein, for example with reference to FIG. 13 .
- the media delivery system 104 retrieves the descriptive search database 610 .
- the media delivery system 104 operates to select one or more media content items 740 ( FIG. 14 ) (also referred to herein as output media content items or the like) based on the relevance score 644 associated with the search keywords 734 .
- An example of the operations 704 and 706 is illustrated and described in more detail herein, for example with reference to FIG. 14 .
- the media delivery system 104 retrieves user profile data 776 ( FIG. 15 ).
- the media delivery system 104 operates to select one or more personalized media content items 772 based on the user profile data 776 .
- An example of the operations 708 and 710 is illustrated and described in more detail herein, for example with reference to FIG. 15 .
- At least one of the operations 702 , 704 , 706 , 708 , and 710 is performed prior to the operation 502 as illustrated in FIG. 5 while the rest of the operations 702 , 704 , 706 , 708 , and 710 follows the operation 502 .
- the operations 702 , 704 , 706 , 708 , and 710 are performed after the operation 502 as illustrated in FIG. 5 .
- FIG. 13 illustrates an example method 730 for analyzing the user query 120 to identify one or more search keywords 734 .
- the method 730 is used to at least partially perform the operation 702 as described in FIG. 12 .
- the user query 120 is a text query that is typed by the user, or a voice request (e.g., an utterance) received through a microphone and converted into a text query.
- the method 730 can be performed at least by a search query process engine 732 that can be included in the media delivery system 104 .
- the search query process engine 732 can operate to analyze the user query 120 and generate the search keywords 734 .
- the user query 120 is tokenized into tokenized terms (e.g., words or phrases), which include descriptors 736 . At least one of the descriptors 736 can be used as the search keywords 734 .
- the user query 120 is a search request “play relaxing jazz for tonight.”
- the search query process engine 732 operates to process the user query 120 and identify the search keywords 734 , such as “relaxing” (or its variants), “tonight” (or its variants), “jazz” (or its variants), etc.
- the search keywords 734 include at least one of the descriptors 736 from the user query 120 , such as descriptive search descriptors 738 and other descriptors 739 .
- the descriptive search descriptors 738 are primarily used to perform a descriptive search as described herein.
- the other descriptors 739 can be used to perform other types of search, such as entity-focused search using media content metadata. As described herein, the other descriptors 739 can be used in addition to the descriptive search descriptors 738 to improve the search results from the descriptive search responsive to the user query 120 .
- the search query process engine 732 refers to the descriptive search database 610 to identify the search keywords 734 by checking to see if any of the tokenized terms (or the descriptors) in the user query 120 match any of the descriptive terms in the descriptive search database 610 . If so, such tokenized terms from the user query 120 are identified as search keywords to be used in the descriptive search.
- FIG. 14 illustrates an example method 750 for selecting media content items 740 based on descriptive search utilizing the descriptive search database 610 .
- the method 750 is used to at least partially perform the operations 704 and 706 as described in FIG. 12 .
- the method 750 can be performed at least by a descriptive search engine 752 that can be included in the media delivery system 104 .
- the descriptive search engine 752 can process the search keywords 734 based on the descriptive search database 610 and return one or more media content items 740 that match the user query 120 .
- the descriptive search engine 752 selects the media content items 740 based on the relevance score 644 .
- the media content items 740 are selected to match the search keywords 734 and have the highest relevance scores 644 .
- a predetermined number of media content items 740 are selected having the highest relevance scores 644 (e.g., top 10 media content items).
- the number of media content items 740 are determined based on other factors, such as a relevance score threshold.
- the descriptive search engine 752 can employ other algorithms to select the media content items 740 in response to the user query 120 .
- the descriptive search engine 752 can operate to obtain a relevance score vector for each of the media content items 234 from the candidate playlists 558 . Further, the descriptive search engine 752 can operate to obtain a corresponding vector for the user query 120 . Then, the descriptive search engine 752 operates to compare each of the relevance score vectors for the media content items 234 with the corresponding vector for the user query 120 , and determine similarity based on the comparison. In some embodiments, cosine similarity can be used to determine the similarity between the vectors. The descriptive search engine 752 can select media content items 740 having relevance score vectors with the highest similarities with the corresponding vector for the user query 120 . Other algorithms are also possible in other embodiments.
- the relevance score 644 for a media content item containing the plurality of search keywords 734 can be computed by multiplication of individual relevance scores 644 for each of the search keywords 734 associated with the media content item. An example of this case is illustrated with the table 620 in FIG. 14 .
- the descriptive search engine 752 can utilize other information about the media content items 234 , such as information from metadata (e.g., the metadata 236 ) of the media content items 234 , to provide even more relevant search results.
- metadata e.g., the metadata 236
- genre information of the media content items 234 are used in view of the other descriptors 739 (e.g., “jazz”) in the search keywords 734 .
- the descriptive search engine 752 can further operate to search for media content items 234 having information from the metadata that matches the other descriptors 739 of the search keywords 734 .
- the metadata of media content items 234 can improve the search results from the descriptive search.
- a search query includes a combination of descriptive search terms and other typical descriptors, such as “play some relaxing jazz music by Miles Davis,” in which case the typical music search algorithm (e.g., using media content item metadata) can be used to identify music by Miles Davis, and the descriptive search algorithm can be used to identify “relaxing” music by the same artist.
- typical music search algorithm e.g., using media content item metadata
- descriptive search algorithm can be used to identify “relaxing” music by the same artist.
- FIG. 15 illustrates an example method 770 for selecting personalized media content items 772 from the media content items 740 .
- the method 770 is used to at least partially perform the operations 708 and 710 as described in FIG. 12 .
- the method 770 can be performed at least by a personalization engine 774 that can be included in the media delivery system 104 .
- the personalization engine 774 can select media content items that are personalized to the user based at least in part on user profile data 776 .
- the method 770 enables the search results to be personalized based on the user profile data 776 .
- the personalization engine 774 utilizes the user profile data 776 to select one or more media content items from the media content items 740 that are selected by the descriptive search as described herein. The personalization engine 774 then outputs the selected media content items as personalized media content items 772 .
- the user profile data 776 include user information.
- the user information can include one or more user specific features that can be used to determine media content personalized to respective users.
- the user specific features can represent the user's preference, media content consumption history, and/or other general features associated with the user of the media playback device 102 .
- a few examples of user specific features include various user media content consumption activities (e.g., listening activities), such as the most listened context (e.g., track, album, artist, playlist, etc.), the most preferred context (e.g., track, album, artist, playlist, etc.), and the most preferred genre or other acoustic feature.
- the user specific features can be obtained based on the user's long-term activities (e.g., over a few months, a year or longer, etc.) or short-term activities (e.g., over a year or less, a month, a week, etc.).
- long-term activities e.g., over a few months, a year or longer, etc.
- short-term activities e.g., over a year or less, a month, a week, etc.
- the user information includes user biographical information and user taste profile information.
- the user biographical information include features relating to personal information of respective users, such as address, age, gender, birth date/year, family, relationships, profession, education, religious preference, sexual preference, association, and other information associated with users.
- the user biographical information can be obtained at least partially from the user account information.
- the user taste profile information include features that can be obtained from taste profiles 242 of respective users, which represent the user's activities with respect to media content consumption, such as user preferences and history of consuming media content, as described herein.
- the user taste profile information includes a user listening history, and other user-related data that can be used to personalize the search results.
- the user taste profile information can further include taste profiles of other users having similar tastes.
- the user profile data 776 can provide scores of one or more user activities with respect to each of the media content items 740 (e.g., “ABC” and “BCD”).
- the user profile data 776 for the particular user provides a playback history score for each media content item 740 to represent how often the user has played the media content item 740 .
- the user profile data 776 can also provide an artist preference score to represent how much the user likes the artist of each of the media content items 740 .
- the user profile data 776 can provide a single score (“Composite Score”) that represents all the scores. In some embodiments, such a single score can be calculated by multiplication of the plurality of scores.
- systems and the methods according to the present disclosure are primarily described to return media content items 740 or 772 as output, it is also possible that the systems and the methods are similarly configured to generate other types of media content, such as media contexts (e.g., playlists, albums, artists, etc.), as output.
- media contexts e.g., playlists, albums, artists, etc.
- an exemplary embodiment of the present disclosure operate to receive an utterance from a voice-enabled device.
- the embodiment can analyze a text version of the utterance to determine one or more descriptors in the utterance.
- the embodiment can optionally sort out a subset of playlists from a corpus of playlists in database based on one or more criteria (e.g., playlist size, time since last modified, various patterns to the playlist title, and other criteria).
- the embodiment can tokenize the playlist title and/or other description into terms.
- the embodiment can tag the media content items in the playlist with the tokenized terms so that the media content items are regarded as keys for the tokenized terms.
- the embodiment can aggregate the tags in each of the media content items in the entire subset of playlists. For example, the embodiment can group by the media content items to collect all the tokenized terms (including duplicates) associated with each of the media content items.
- the embodiment can apply term frequency and inverse document frequency (TFIDF) to weight the tokenized terms among the different media content items.
- TDF term frequency and inverse document frequency
- the embodiment can generate an inverted index that maps the tokenized terms to the media content items.
- the embodiment can perform search for media content items using the inverted index based on the descriptors in the utterance.
- the search can be performed by comparing the descriptors in the utterance with the tokenized terms in the inverted index, and then determining one or more of the tokenized terms in the inverted index and identifying the media content items associated with the determined tokenized terms.
- the embodiment can then return the identified media content items to the voice-enabled device for playback or other controlling of playback.
Abstract
A descriptive media content search solution is provided to allow a user to search for media content that better matches a user's descriptive search request. The descriptive media content search solution utilizes an extensive catalog of playlists each having a playlist description, such as a playlist title or other descriptive text, and identifies additional descriptive information for media content items to be searched. The descriptive media content search solution can set up a descriptive search database and utilize the descriptive search database to conduct a descriptive search responsive to the user's descriptive search request.
Description
- This application is a continuation of U.S. application Ser. No. 17/489,937, filed Sep. 30, 2021, which is a continuation of U.S. application Ser. No. 15/988,943, filed on May 24, 2018, titled DESCRIPTIVE MEDIA CONTENT SEARCH, the disclosures of which are hereby incorporated by reference in their entireties. To the extent appropriate a claim of priority is made to each of the above-disclosed applications.
- Music search algorithms are typically based on known metadata of the music that is stored in a music library metadata database. The metadata includes typical information that is commonly associated with music, such as the name of the song, the name of the artist, a genre, a release date, and the like. This type of music search may be helpful for a user to identify a particular track if the user knows that information about the track. However, the music search of this type does not provide relevant search results if the user query describes something that is not included in the music metadata database.
- In general terms, the present disclosure is directed to descriptive media content search. In one possible configuration and by non-limiting example, the descriptive media content search utilizes playlist descriptions to provide additional information to describe media content. Various aspects are described in this disclosure, which include, but are not limited to, the following aspects.
- One aspect is a method of selecting and providing media content responsive to a query. The method includes identifying descriptive terms associated with media content items, and selecting and providing one or more media content items responsive to a query. The descriptive terms associated with media content items may be identified by identifying a plurality of playlists, each of the plurality of playlists associated with one or more media content items and having a playlist description; and for each of the plurality of playlists, tokenizing the playlist description into one or more tokenized terms; and associating each of the media content items in the respective playlist with the tokenized terms. The media content items responsive to a query may be selected and provided by receiving the query including at least one descriptor, identifying the descriptor from the query, and identifying one or more media content items, from the plurality of playlists, associated with the descriptor based on the tokenized terms.
- In certain examples, the user query is a voice request from a user. In certain examples, the method may include converting the voice request into text data, determining an intent and at least one slot from the text data, and identifying the descriptor from at least one of the intent and the at least one slot.
- In certain examples, the descriptive terms associated with media content items may be identified by, for each media content item in the plurality of playlists, weighting the media content items associated with the tokenized terms. The identification of the media content items may be based on the tokenized terms and the weightings of the tokenized terms.
- In certain examples, the method may include calculating a first term frequency, the first term frequency associated with a frequency of each of the tokenized terms appearing in each of the media content items; calculating a second term frequency, the second term frequency associated with a frequency of each of the tokenized terms appearing in all of the media content items; and calculating a relevance score based on the first term frequency and the second term frequency, the relevance score representative of relevancy of each of the tokenized terms to each of the media content items.
- In certain examples, the media content items are identified from the plurality of playlists based at least in part on the relevance score.
- In certain examples, weighting is performed using term frequency and inverse document frequency (TFIDF).
- In certain examples, the method may include, after weighting the media content items associated with the tokenized terms, generating an inverted index that stores the associations between the media content items with the tokenized terms.
- In certain examples, the playlist description includes a playlist title and a descriptive text. In certain examples, the descriptive text is obtained from at least one of user reviews, expert reviews, and statistics.
- In certain examples, the method may include retrieving a user profile associated with a user who provided the user query, and determining personalized media content items from the resulting media content items based on the user profile.
- In certain examples, the method may include retrieving a descriptive term lexicon. The playlist description may be tokenized into one or more tokenized terms based on the descriptive term lexicon.
- In certain examples, the method may include retrieving metadata of media content items of the playlists. The media content items may be identified from the plurality of playlists based in part on the metadata.
- Another aspect is a media delivery system for selecting and providing media content responsive to a user query. The media delivery system includes at least one processing device and at least one data storage device storing data instructions executable by the at least one processing device to cause the media delivery system to identify descriptive terms associated with media content items, and select and provide one or more media content items responsive to a user query. The descriptive terms associated with media content items may be identified by identifying a plurality of playlists, each of the plurality of playlists associated with one or more media content items and having a playlist description; and, for each of the plurality of playlists, tokenizing the playlist description into one or more tokenized terms; and associating each of the media content items in the respective playlist with the tokenized terms. The media content items responsive to a user query may be selected and provided by receiving the user query including at least one descriptor, identifying the descriptor from the user query, and identifying one or more media content items, from the plurality of playlists, associated with the descriptor based on the tokenized terms.
- In certain examples, the data instructions may further cause the media delivery system to, for each media content item in the plurality of playlists, weight the media content items associated with the tokenized terms. the media content items may be identified based on the tokenized terms and the weightings of the tokenized terms.
- In certain examples, the data instructions may further cause the media delivery system to calculate a first term frequency, the first term frequency associated with a frequency of each of the tokenized terms appearing in each of the media content items; calculate a second term frequency, the second term frequency associated with a frequency of each of the tokenized terms appearing in all of the media content items; and calculate a relevance score based on the first term frequency and the second term frequency, the relevance score representative of relevancy of each of the tokenized terms to each of the media content items.
- In certain examples, the media content items may be identified from the plurality of playlists based at least in part on the relevance score.
- In certain examples, the data instructions may further cause the media delivery system to weight the media content items associated with the tokenized terms using term frequency and inverse document frequency (TFIDF).
- In certain examples, the data instructions may further cause the media delivery system to, after weighting the media content items associated with the tokenized terms, generate an inverted index that stores the associations between the media content items with the tokenized terms.
- In certain examples, the playlist description may include a playlist title and a descriptive text.
-
FIG. 1 illustrates an example system for providing media content to a user. -
FIG. 2 is a block diagram of an example media playback device of the system shown inFIG. 1 . -
FIG. 3 is a flowchart of an example method for selecting and providing media content responsive to a user query. -
FIG. 4 is a flowchart of an example method for processing an utterance of the user query. -
FIG. 5 is a flowchart of an example method for performing descriptive search for media content based on the user query. -
FIG. 6 is a flowchart of an example method for generating and updating descriptive search database. -
FIG. 7 illustrates an example method for determining candidate playlists. -
FIG. 8 illustrates an example method for identifying one or more descriptive terms from the candidate playlists. -
FIG. 9 illustrates an example data structure that associates the descriptive terms with media content items. -
FIG. 10 is a flowchart of an example method for weighting the media content items associated with the descriptive terms. -
FIG. 11 illustrates an example descriptive search database. -
FIG. 12 is a flowchart of an example method for performing a descriptive search from the descriptive search database. -
FIG. 13 illustrates an example method for analyzing the user query to identify one or more search keywords. -
FIG. 14 illustrates an example method for selecting media content items based on descriptive search utilizing the descriptive search database. -
FIG. 15 illustrates an example method for selecting personalized media content items. - Various embodiments will be described in detail with reference to the drawings, wherein like reference numerals represent like parts and assemblies throughout the several views.
- In general, the present disclosure relates to a descriptive media content search solution that allows for searching for and identifying media content based on descriptive terms.
- In various media content search contexts, a search request can include a series of descriptors. Such a search request with descriptors may not return satisfactory search results if used with typical entity-focused search indexes, such as search algorithms based on known metadata of the media content that is stored in a music library metadata database. For example, when the search request includes descriptors that are not found in the metadata, the search results do not fully reflect the descriptors in the search request. By way of example, when a search query is “play some relaxing jazz music,” a descriptive term to be searched can include “relaxing” (or its variant) or “relaxing jazz” (or its variant), and this descriptive term is not typically found in the metadata of media content.
- The present disclosure provides a solution to allow a user to search for media content that better matches the user's descriptive search request. Further, the solution according to the present disclosure can provide relevant search results even when the user do not know specific information about the media content the user wants to listen to, but can describe general characteristic of the media content with descriptive terms. Examples of descriptive terms relate to moods (relaxing, happy, sad, somber, etc.), types of activities (such as running, working out, driving, dancing, etc.), or any other descriptors that are not typically found in the metadata of media content but describe some aspect of the media content.
- An exemplary embodiment of the present disclosure utilizes an extensive catalog of playlists each having a playlist description, such as a playlist title or other descriptive text, and identifies additional descriptive information for media content items to be searched. The embodiment of the present disclosure can involve setting up a descriptive search database, and then utilizing the descriptive search database to conduct a descriptive search responsive to a descriptive search query. The embodiment of the present disclosure can analyze a variety of data associated with the catalog of playlists, such as the playlist descriptions, to generate descriptive terms (e.g., tokens) for media content to be searched for. The descriptive search database can be generated with the descriptive terms that map to media content items in the catalog of playlists.
-
FIG. 1 illustrates anexample system 100 for providing media content to a user U. Thesystem 100 includes amedia playback device 102, amedia delivery system 104, and adata communication network 106. Themedia delivery system 104 includes a descriptive mediacontent search engine 110. Anexample user query 120 and amedia output 122 are also shown. - The
media playback device 102 operates to provide media content to a user U. As described herein, themedia playback device 102 operates to receive theuser query 120 and provide themedia output 122 to the user U according to theuser query 120. In some embodiments, themedia playback device 102 operates to communicate with a system external to themedia playback device 102, such as themedia delivery system 104. Themedia playback device 102 can interact with themedia delivery system 104 to process theuser query 120 and identify media content in response to theuser query 120. In some embodiments, themedia playback device 102 operates to receive the media content that is identified and provided (e.g., streamed, transmitted, etc.) by themedia delivery system 104. In some embodiments, themedia playback device 102 operates to play the media content and generate themedia output 122 using a media output device (e.g., a speaker) therein. In other embodiments, themedia playback device 102 operates to transmit the media content to another device for playback, such as an external speaker or another media playback device (e.g., a vehicle entertainment system or a home entertainment system). An example of themedia playback device 102 is illustrated and described in more detail herein, such as with reference toFIG. 2 . - The
media delivery system 104 operates to provide media content to one or more media playback devices, such as themedia playback device 102, via thenetwork 106. An example of themedia delivery system 104 is illustrated and described in further detail herein, such as with reference toFIG. 2 . - The
network 106 is a data communication network that facilitates data communication between themedia playback device 102 and themedia delivery system 104. Thenetwork 106 typically includes a set of computing devices and communication links between the computing devices. The computing devices in thenetwork 106 use the links to enable communication among the computing devices in the network. Thenetwork 106 can include one or more routers, switches, mobile access points, bridges, hubs, intrusion detection devices, storage devices, standalone server devices, blade server devices, sensors, desktop computers, firewall devices, laptop computers, handheld computers, mobile telephones, vehicular computing devices, and other types of computing devices. - In various embodiments, the
network 106 includes various types of communication links. For example, thenetwork 106 can include wired and/or wireless links, including cellular, Bluetooth®, Wi-Fi®, ultra-wideband (UWB), 802.11, ZigBee, near field communication (NFC), an ultrasonic data transmission, and other types of wireless links. Furthermore, in various embodiments, thenetwork 106 is implemented at various scales. For example, thenetwork 106 can be implemented as one or more vehicle area networks, local area networks (LANs), metropolitan area networks, subnets, wide area networks (WAN) (such as the Internet), or can be implemented at another scale. Further, in some embodiments, thenetwork 106 includes multiple networks, which may be of the same type or of multiple different types. - Referring still to
FIG. 1 , in some embodiments, themedia delivery system 104 includes the descriptive mediacontent search engine 110 that operates to determine media content in response to theuser query 120. - As described herein, the
user query 120 includes a search request from the user U to identify media content. Theuser query 120 can be also referred to herein as a search query, a search request, or the like. In some embodiments, theuser query 120 can be a text that is typed using themedia playback device 102 or another computing device. In other embodiments, theuser query 120 can be a voice request received through a sound detection device (e.g., a microphone). As described herein, the voice request can be processed into a text query suitable for media content search. - In some embodiments, the
user query 120 includes one or more descriptive terms (also referred to herein as descriptors) that can be used as one or more keywords to identify media content associated with the keywords. In the illustrated example ofFIG. 1 , theuser query 120 can be a voice request that is converted into a text query of “play relaxing jazz for tonight.” In this example, the descriptive terms in theuser query 120 can include “relaxing” or its variant (e.g., “relax” or “relaxation”), “tonight” or its variant (e.g., “night”), and/or “jazz” or its variant (e.g., “jazzy”). Alternatively, in this example, the descriptive terms in theuser query 120 can include “relaxing jazz” as a single phrase instead of “relaxing” and “jazz” as two separate words. Other examples are also possible. - The descriptive media
content search engine 110 operates to select media content based on one or more descriptive terms in theuser query 120. For example, the descriptive mediacontent search engine 110 can identify media content associated with at least one of the descriptive terms of theuser query 120, such as “relaxing,” “tonight,” “relaxing jazz,” or “jazz” in the illustrated example. An example of the descriptive mediacontent search engine 110 is illustrated and described in more detail herein, such as with reference toFIG. 3 . -
FIG. 2 is a block diagram of an example embodiment of themedia playback device 102 of the mediacontent provision system 100 shown inFIG. 1 . In this example, themedia playback device 102 includes a user input device 130, adisplay device 132, a wirelessdata communication device 134, a mediacontent output device 140, aprocessing device 148, and amemory device 150. - The
media playback device 102 operates to play media content. For example, themedia playback device 102 is configured to play media content that is provided (e.g., streamed or transmitted) by a system external to themedia playback device 102, such as themedia delivery system 104, another system, or a peer device. In other examples, themedia playback device 102 operates to play media content stored locally on themedia playback device 102. In yet other examples, themedia playback device 102 operates to play media content that is stored locally as well as media content provided by other systems. - In some embodiments, the
media playback device 102 is a handheld or portable entertainment device, smartphone, tablet, watch, wearable device, or any other type of computing device capable of playing media content. In other embodiments, themedia playback device 102 is a laptop computer, desktop computer, television, gaming console, set-top box, network appliance, blue-ray or DVD player, media player, stereo, or radio. - In some embodiments, the
media playback device 102 is a system dedicated for streaming personalized media content in a vehicle environment. An example of such a vehicular media playback device is described in U.S. patent application Ser. No. 15/973,198, titled PERSONAL MEDIA STREAMING APPLIANCE SYSTEM, filed May 7, 2018, the disclosure of which is hereby incorporated by reference in its entirety. - The user input device 130 operates to receive a
user input 152 from a user U for controlling themedia playback device 102. As illustrated, theuser input 152 can include a manual input 154 and avoice input 156. In some embodiments, the user input device 130 includes amanual input device 160 and asound detection device 162. - The
manual input device 160 operates to receive the manual input 154 for controlling playback of media content via themedia playback device 102. In some embodiments, themanual input device 160 includes one or more buttons, keys, touch levers, switches, and/or other mechanical input devices for receiving the manual input 154. For example, themanual input device 160 includes a text entry interface, such as a mechanical keyboard, a virtual keyboard, or a handwriting input device, which is configured to receive a text input, such as a text version of theuser query 120. In addition, in some embodiments, the manual input 154 is received for managing various pieces of information transmitted via themedia playback device 102 and/or controlling other functions or aspects associated with themedia playback device 102. - The
sound detection device 162 operates to detect and record sounds from proximate themedia playback device 102. For example, thesound detection device 162 can detect sounds including thevoice input 156. In some embodiments, thesound detection device 162 includes one or more acoustic sensors configured to detect sounds proximate themedia playback device 102. For example, acoustic sensors of thesound detection device 162 include one or more microphones. Various types of microphones can be used for thesound detection device 162 of themedia playback device 102. - In some embodiments, the
voice input 156 is a user's voice (also referred to herein as an utterance) for controlling playback of media content via themedia playback device 102. For example, thevoice input 156 includes a voice version of theuser query 120 received from thesound detection device 162 of themedia playback device 102. In addition, thevoice input 156 is a user's voice for managing various data transmitted via themedia playback device 102 and/or controlling other functions or aspects associated with themedia playback device 102. - In some embodiments, the sounds detected by the
sound detection device 162 can be processed by thesound processing engine 180 of themedia playback device 102 as described below. - Referring still to
FIG. 2 , thedisplay device 132 operates to display information to the user U. Examples of such information include media content playback information, notifications, and other information. In some embodiments, thedisplay screen 132 is configured as a touch sensitive display and includes themanual input device 160 of the user input device 130 for receiving the manual input 154 from a selector (e.g., a finger, stylus etc.) controlled by the user U. In some embodiments, therefore, thedisplay screen 132 operates as both a display device and a user input device. The touchsensitive display screen 132 operates to detect inputs based on one or both of touches and near-touches. In some embodiments, thedisplay screen 132 displays a graphical user interface for interacting with themedia playback device 102. Other embodiments of thedisplay screen 132 do not include a touch sensitive display screen. Some embodiments include a display device and one or more separate user interface devices. Further, some embodiments do not include a display device. - The
data communication device 134 operates to enable themedia playback device 102 to communicate with one or more computing devices over one or more networks, such as thenetwork 106 For example, thedata communication device 134 is configured to communicate with themedia delivery system 104 and receive media content from themedia delivery system 104 at least partially via thenetwork 106. Thedata communication device 134 can be a network interface of various types which connects themedia playback device 102 to thenetwork 106. Examples of thedata communication device 134 include wired network interfaces and wireless network interfaces. Wireless network interfaces includes infrared, BLUETOOTH® wireless technology, 802.11a/b/g/n/ac, and cellular or other radio frequency interfaces in at least some possible embodiments. Examples of cellular network technologies include LTE, WiMAX, UMTS, CDMA2000, GSM, cellular digital packet data (CDPD), and Mobitex. - The media
content output device 140 operates to output media content. In some embodiments, the mediacontent output device 140 generates themedia output 122 for the user U. In some embodiments, the mediacontent output device 140 includes one or more embeddedspeakers 164 which are incorporated in themedia playback device 102. - Alternatively or in addition, some embodiments of the
media playback device 102 include anexternal speaker interface 166 as an alternative output of media content. Theexternal speaker interface 166 is configured to connect themedia playback device 102 to another system having one or more speakers, such as headphones, a portal speaker, and a vehicle entertainment system, so that themedia output 122 is generated via the speakers of the other system external to themedia playback device 102. Examples of theexternal speaker interface 166 include an audio output jack, a USB port, a Bluetooth transmitter, a display panel, and a video output jack. Other embodiments are possible as well. For example, theexternal speaker interface 166 is configured to transmit a signal that can be used to reproduce an audio signal by a connected or paired device such as headphones or a speaker. - The
processing device 148, in some embodiments, comprises one or more central processing units (CPU). In other embodiments, theprocessing device 148 additionally or alternatively includes one or more digital signal processors, field-programmable gate arrays, or other electronic circuits. - The
memory device 150 typically includes at least some form of computer-readable media. Thememory device 150 can include at least one data storage device. Computer readable media includes any available media that can be accessed by themedia playback device 102. By way of example, computer-readable media includes computer readable storage media and computer readable communication media. - Computer readable storage media includes volatile and nonvolatile, removable and non-removable media implemented in any device configured to store information such as computer readable instructions, data structures, program modules, or other data. Computer readable storage media includes, but is not limited to, random access memory, read only memory, electrically erasable programmable read only memory, flash memory and other memory technology, compact disc read only memory, blue ray discs, digital versatile discs or other optical storage, magnetic storage devices, or any other medium that can be used to store the desired information and that can be accessed by the
media playback device 102. In some embodiments, computer readable storage media is non-transitory computer readable storage media. - Computer readable communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media. The term “modulated data signal” refers to a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, computer readable communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, radio frequency, infrared, and other wireless media. Combinations of any of the above are also included within the scope of computer readable media.
- The
memory device 150 operates to store data and instructions. In some embodiments, thememory device 150 stores instructions for amedia content cache 172, acaching management engine 174, amedia playback engine 176, asound processing engine 180, and avoice interaction engine 182. - Some embodiments of the
memory device 150 include themedia content cache 172. Themedia content cache 172 stores media content items, such as media content items that have been received from themedia delivery system 104. The media content items stored in themedia content cache 172 may be stored in an encrypted or unencrypted format. In some embodiments, themedia content cache 172 also stores metadata about media content items such as title, artist name, album name, length, genre, mood, era, etc. Themedia content cache 172 can further store playback information about the media content items and/or other information associated with the media content items. - The
caching management engine 174 is configured to receive and cache media content in themedia content cache 172 and manage the media content stored in themedia content cache 172. In some embodiments, when media content is streamed from themedia delivery system 104, thecaching management engine 174 operates to cache at least a portion of the media content into themedia content cache 172. In other embodiments, thecaching management engine 174 operates to cache at least a portion of media content into themedia content cache 172 while online so that the cached media content is retrieved for playback while themedia playback device 102 is offline. - The
media playback engine 176 operates to play media content to the user U. As described herein, themedia playback engine 176 is configured to communicate with themedia delivery system 104 to receive one or more media content items (e.g., through the media stream 232). In other embodiments, themedia playback engine 176 is configured to play media content that is locally stored in themedia playback device 102. - In some embodiments, the
media playback engine 176 operates to retrieve one or more media content items that are either locally stored in themedia playback device 102 or remotely stored in themedia delivery system 104. In some embodiments, themedia playback engine 176 is configured to send a request to themedia delivery system 104 for media content items and receive information about such media content items for playback. - The
sound processing engine 180 is configured to receive sound signals obtained from thesound detection device 162 and process the sound signals to identify different sources of the sounds received via thesound detection device 162. In some embodiments, thesound processing engine 180 operates to filter the user's voice input 156 (e.g., a voice request of the user query 120) from noises included in the detected sounds. Various noise cancellation technologies, such as active noise control or cancelling technologies or passive noise control or cancelling technologies, can be used to filter the voice input from ambient noise. In examples, thesound processing engine 180 filters out omni-directional noise and preserves directional noise (e.g., an audio input difference between two microphones) in audio input. In examples, thesound processing engine 180 removes frequencies above or below human speaking voice frequencies. In examples, thesound processing engine 180 subtracts audio output of the device from the audio input to filter out the audio content being provided by the device. (e.g., to reduce the need of the user to shout over playing music). In examples, thesound processing engine 180 performs echo cancellation. By using one or more of these techniques, thesound processing engine 180 provides sound processing customized for use in a vehicle environment. - In other embodiments, the
sound processing engine 180 operates to process the received sound signals to identify the sources of particular sounds of the sound signals, such as people's conversation in the vehicle, the vehicle engine sound, or other ambient sounds associated with the vehicle. - In some embodiments, a recording of sounds captured using the
sound detection device 162 can be analyzed using speech recognition technology to identify words spoken by the user. The words may be recognized as commands from the user that alter the playback of media content and/or other functions or aspects of themedia playback device 102. In some embodiments, the words and/or the recordings may also be analyzed using natural language processing and/or intent recognition technology to determine appropriate actions to take based on the spoken words. Additionally or alternatively, thesound processing engine 180 may determine various sound properties about the sounds proximate themedia playback device 102 such as volume, dominant frequency or frequencies, etc. These sound properties may be used to make inferences about the environment proximate to themedia playback device 102. - The
voice interaction engine 182 operates to cooperate with the media delivery system 104 (e.g., avoice interaction server 204 thereof) to identify a command (e.g., a user intent) that is conveyed by thevoice input 156. In some embodiments, thevoice interaction engine 182 transmits the user'svoice input 156 that is detected by thesound processing engine 180 to themedia delivery system 104 so that themedia delivery system 104 operates to determine a command intended by thevoice input 156. In other embodiments, at least some of the determination process of the command can be performed locally by thevoice interaction engine 182. - In addition, some embodiments of the
voice interaction engine 182 can operate to cooperate with the media delivery system 104 (e.g., thevoice interaction server 204 thereof) to provide a voice assistant that performs various voice-based interactions with the user, such as voice feedbacks, voice notifications, voice recommendations, and other voice-related interactions and services. - Referring still to
FIG. 2 , themedia delivery system 104 includes a media content server 200, a mediacontent search server 202, avoice interaction server 204, and a user command interpretation server 206. - The
media delivery system 104 comprises one or more computing devices and provides media content to themedia playback device 102 and, in some embodiments, other media playback devices as well. In addition, themedia delivery system 104 interacts with themedia playback device 102 to provide themedia playback device 102 with various functionalities. - In at least some embodiments, the media content server 200, the media
content search server 202, thevoice interaction server 204, and the user command interpretation server 206 are provided by separate computing devices. In other embodiments, the media content server 200, the mediacontent search server 202, thevoice interaction server 204, and the user command interpretation server 206 are provided by the same computing device(s). Further, in some embodiments, at least one of the media content server 200, the mediacontent search server 202, thevoice interaction server 204, and the user command interpretation server 206 is provided by multiple computing devices. For example, the media content server 200, the mediacontent search server 202, thevoice interaction server 204, and the user command interpretation server 206 may be provided by multiple redundant servers located in multiple geographic locations. - Although
FIG. 2 shows a single media content server 200, a single mediacontent search server 202, a singlevoice interaction server 204, and a single user command interpretation server 206, some embodiments include multiple media content servers, media content search servers, voice interaction servers, and user command interpretation servers. In these embodiments, each of the multiple media content servers, media content search servers, voice interaction servers, and user command interpretation servers may be identical or similar to the media content server 200, the mediacontent search server 202, thevoice interaction server 204, and the user command interpretation server 206, respectively, as described herein, and may provide similar functionality with, for example, greater capacity and redundancy and/or services from multiple geographic locations. Alternatively, in these embodiments, some of the multiple media content servers, the media content search servers, the voice interaction servers, and/or the user command interpretation servers may perform specialized functions to provide specialized services. Various combinations thereof are possible as well. - The media content server 200 transmits stream media to media playback devices such as the
media playback device 102. In some embodiments, the media content server 200 includes amedia server application 212, aprocessing device 214, amemory device 216, and anetwork access device 218. Theprocessing device 214 and thememory device 216 may be similar to theprocessing device 148 and thememory device 150, respectively, which have each been previously described. Therefore, the description of theprocessing device 214 and thememory device 216 are omitted for brevity purposes. - The
network access device 218 operates to communicate with other computing devices over one or more networks, such as thenetwork 106. Examples of the network access device include one or more wired network interfaces and wireless network interfaces. Examples of such wireless network interfaces of thenetwork access device 218 include wireless wide area network (WWAN) interfaces (including cellular networks) and wireless local area network (WLANs) interfaces. In other examples, other types of wireless interfaces can be used for thenetwork access device 218. - In some embodiments, the
media server application 212 is configured to stream media content, such as music or other audio, video, or other suitable forms of media content. Themedia server application 212 includes amedia stream service 222, amedia application interface 224, and amedia data store 226. Themedia stream service 222 operates to buffer media content, such asmedia content items more media streams - The
media application interface 224 can receive requests or other communication from media playback devices or other systems, such as themedia playback device 102, to retrieve media content items from the media content server 200. For example, inFIG. 2 , themedia application interface 224 receives communication from themedia playback device 102 to receive media content from the media content server 200. - In some embodiments, the
media data store 226 storesmedia content items 234,media content metadata 236, media contexts 238, user accounts 240, and taste profiles 242. Themedia data store 226 may comprise one or more databases and file systems. Other embodiments are possible as well. - As discussed herein, the media content items 234 (including the media content items 230) may be audio, video, or any other type of media content, which may be stored in any format for storing media content.
- The
media content metadata 236 provides various information associated with themedia content items 234. In addition or alternatively, themedia content metadata 236 provides various information associated with the media contexts 238. In some embodiments, themedia content metadata 236 includes one or more of title, artist name, album name, length, genre, mood, era, etc. - The
media content metadata 236 operates to provide various pieces of information associated with themedia content items 234 and/or the media contexts 238. In some embodiments, themedia content metadata 236 includes one or more of title, artist name, album name, length, genre, mood, era, etc. - In some embodiments, the
media content metadata 236 includes acoustic metadata, cultural metadata, and explicit metadata. The acoustic metadata may be derived from analysis of the track and refers to a numerical or mathematical representation of the sound of a track. Acoustic metadata may include temporal information such as tempo, rhythm, beats, downbeats, tatums, patterns, sections, or other structures. Acoustic metadata may also include spectral information such as melody, pitch, harmony, timbre, chroma, loudness, vocalness, or other possible features. Acoustic metadata may take the form of one or more vectors, matrices, lists, tables, and other data structures. Acoustic metadata may be derived from analysis of the music signal. One form of acoustic metadata, commonly termed an acoustic fingerprint, may uniquely identify a specific track. Other forms of acoustic metadata may be formed by compressing the content of a track while retaining some or all of its musical characteristics. - The cultural metadata refers to text-based information describing listeners' reactions to a track or song, such as styles, genres, moods, themes, similar artists and/or songs, rankings, etc. Cultural metadata may be derived from expert opinion such as music reviews or classification of music into genres. Cultural metadata may be derived from listeners through websites, chatrooms, blogs, surveys, and the like. Cultural metadata may include sales data, shared collections, lists of favorite songs, and any text information that may be used to describe, rank, or interpret music. Cultural metadata may also be generated by a community of listeners and automatically retrieved from Internet sites, chat rooms, blogs, and the like. Cultural metadata may take the form of one or more vectors, matrices, lists, tables, and other data structures. A form of cultural metadata particularly useful for comparing music is a description vector. A description vector is a multi-dimensional vector associated with a track, album, or artist. Each term of the description vector indicates the probability that a corresponding word or phrase would be used to describe the associated track, album or artist.
- The explicit metadata refers to factual or explicit information relating to music. Explicit metadata may include album and song titles, artist and composer names, other credits, album cover art, publisher name and product number, and other information. Explicit metadata is generally not derived from the music itself or from the reactions or opinions of listeners.
- At least some of the
metadata 236, such as explicit metadata (names, credits, product numbers, etc.) and cultural metadata (styles, genres, moods, themes, similar artists and/or songs, rankings, etc.), for a large library of songs or tracks can be evaluated and provided by one or more third party service providers. Acoustic and cultural metadata may take the form of parameters, lists, matrices, vectors, and other data structures. Acoustic and cultural metadata may be stored as XML files, for example, or any other appropriate file type. Explicit metadata may include numerical, text, pictorial, and other information. Explicit metadata may also be stored in an XML, or other file. All or portions of the metadata may be stored in separate files associated with specific tracks. All or portions of the metadata, such as acoustic fingerprints and/or description vectors, may be stored in a searchable data structure, such as a k-D tree or other database format. - Referring still to
FIG. 2 , each of the media contexts 238 is used to identify one or moremedia content items 234. In some embodiments, the media contexts 238 are configured to group one or moremedia content items 234 and provide a particular context to the group ofmedia content items 234. Some examples of the media contexts 238 include albums, artists, playlists, and individual media content items. By way of example, where a media context 238 is an album, the media context 238 can represent that themedia content items 234 identified by the media context 238 are associated with that album. - As described above, the media contexts 238 can include playlists 239. The playlists 239 are used to identify one or more of the
media content items 234. In some embodiments, the playlists 239 identify a group of themedia content items 234 in a particular order. In other embodiments, the playlists 239 merely identify a group of themedia content items 234 without specifying a particular order. Some, but not necessarily all, of themedia content items 234 included in a particular one of the playlists 239 are associated with a common characteristic such as a common genre, mood, or era. - In some embodiments, a user can listen to media content items in a playlist 239 by selecting the playlist 239 via a media playback device, such as the
media playback device 102. The media playback device then operates to communicate with themedia delivery system 104 so that themedia delivery system 104 retrieves the media content items identified by the playlist 239 and transmits data for the media content items to the media playback device for playback. - In some embodiments, the playlist 239 includes one or more playlist descriptions, such as a
playlist description 570 as illustrated inFIG. 8 . The playlist descriptions include information associated with the playlist 239. The playlist descriptions can include a playlist title, such as aplaylist title 574 as illustrated inFIG. 8 . The playlist title is a title of the playlist. In some embodiments, the playlist title can be provided by a user using themedia playback device 102. In other embodiments, the playlist title can be provided by a media content provider (or a media-streaming service provider). In yet other embodiments, the playlist title can be automatically generated. - Other examples of playlist descriptions include a descriptive text, such as a
descriptive text 576 as illustrated inFIG. 8 . The descriptive text can be provided by the user and/or the media content provider, which is to represent the corresponding playlist 239. In other embodiments, the descriptive text of the playlist description can be obtained from one or more other sources. Such other sources can include expert opinion (e.g., music reviews or classification of music into genres), user opinion (e.g., reviews through websites, chatrooms, blogs, surveys, and the like), statistics (e.g., sales data), shared collections, lists of favorite playlists, and any text information that may be used to describe, rank, or interpret the playlist or music associated with the playlist. In some embodiments, the playlist descriptions can also be generated by a community of listeners and automatically retrieved from Internet sites, chat rooms, blogs, and the like. - In some embodiments, the playlist descriptions can take the form of one or more vectors, matrices, lists, tables, and other data structures. A form of cultural metadata particularly useful for comparing music is a description vector. A description vector is a multi-dimensional vector associated with a track, album, or artist. Each term of the description vector indicates the probability that a corresponding word or phrase would be used to describe the associated track, album or artist. Each term of the description vector indicates the probability that a corresponding word or phrase would be used to describe the associated track, album or artist.
- In some embodiments, the playlist 239 includes a list of media content item identifications (IDs), such as media content item identifications (IDs) 578 as illustrated in
FIG. 8 . The list of media content item identifications includes one or more media content item identifications that refer to respectivemedia content items 234. Each media content item is identified by a media content item ID and includes various pieces of information, such as a media content item title, artist identification (e.g., individual artist name or group name, or multiple artist names or group names), and media content item data. In some embodiments, the media content item title and the artist ID are part of themedia content metadata 236, which can further include other attributes of the media content item, such as album name, length, genre, mood, era, etc. as described herein. - At least some of the playlists 239 may include user-created playlists. For example, a user of a media streaming service provided using the
media delivery system 104 can create a playlist 239 and edit the playlist 239 by adding, removing, and rearranging media content items in the playlist 239. A playlist 239 can be created and/or edited by a group of users together to make it a collaborative playlist. In some embodiments, user-created playlists can be available to a particular user only, a group of users, or to the public based on a user-definable privacy setting. - In some embodiments, when a playlist is created by a user or a group of users, the
media delivery system 104 operates to generate a list of media content items recommended for the particular user or the particular group of users. In some embodiments, such recommended media content items can be selected based at least on the taste profiles 242 as described herein. Other information or factors can be used to determine the recommended media content items. Examples of determining recommended media content items are described in U.S. patent application Ser. No. 15/858,377, titled MEDIA CONTENT ITEM RECOMMENDATION SYSTEM, filed Dec. 29, 2017, the disclosure of which is hereby incorporated by reference in its entirety. - In addition or alternatively, at least some of the playlists 239 are created by a media streaming service provider. For example, such provider-created playlists can be automatically created by the
media delivery system 104. In some embodiments, a provider-created playlist can be customized to a particular user or a particular group of users. By way of example, a playlist for a particular user can be automatically created by themedia delivery system 104 based on the user's listening history (e.g., the user's taste profile) and/or listening history of other users with similar tastes. In other embodiments, a provider-created playlist can be configured to be available for the public in general. Provider-created playlists can also be sharable with other users. - The user accounts 240 are used to identify users of a media streaming service provided by the
media delivery system 104. In some embodiments, a user account 240 allows a user to authenticate to themedia delivery system 104 and enable the user to access resources (e.g., media content items, playlists, etc.) provided by themedia delivery system 104. In some embodiments, the user can use different devices to log into the user account and access data associated with the user account in themedia delivery system 104. User authentication information, such as a username, an email account information, a password, and other credentials, can be used for the user to log into his or her user account. It is noted that, where user data is to be protected, the user data is handled according to robust privacy and data protection policies and technologies. For instance, whenever personally identifiable information and any other information associated with users is collected and stored, such information is managed and secured using security measures appropriate for the sensitivity of the data. Further, users can be provided with appropriate notice and control over how any such information is collected, shared, and used. - The taste profiles 242 contain records indicating media content tastes of users. A taste profile can be associated with a user and used to maintain an in-depth understanding of the music activity and preference of that user, enabling personalized recommendations, taste profiling and a wide range of social music applications. Libraries and wrappers can be accessed to create taste profiles from a media library of the user, social website activity and other specialized databases to obtain music preferences.
- In some embodiments, each
taste profile 242 is a representation of musical activities, such as user preferences and historical information about the users' consumption of media content, and can include a wide range of information such as artist plays, song plays, skips, dates of listen by the user, songs per day, playlists, play counts, start/stop/skip data for portions of a song or album, contents of collections, user rankings, preferences, or other mentions received via a client device, or other media plays, such as web sites visited, book titles, movies watched, playing activity during a movie or other presentations, ratings, or terms corresponding to the media, such as “comedy,” etc. - In addition, the taste profiles 242 can include other information. For example, the taste profiles 242 can include libraries and/or playlists of media content items associated with the user. The taste profiles 242 can also include information about the user's relationships with other users (e.g., associations between users that are stored by the
media delivery system 104 or on a separate social media site). - The taste profiles 242 can be used for a number of purposes. One use of taste profiles is for creating personalized playlists (e.g., personal playlisting). An API call associated with personal playlisting can be used to return a playlist customized to a particular user. For example, the media content items listed in the created playlist are constrained to the media content items in a taste profile associated with the particular user. Another example use case is for event recommendation. A taste profile can be created, for example, for a festival that contains all the artists in the festival. Music recommendations can be constrained to artists in the taste profile. Yet another use case is for personalized recommendation, where the contents of a taste profile are used to represent an individual's taste. This API call uses a taste profile as a seed for obtaining recommendations or playlists of similar artists. Yet another example of taste profile use case is referred to as bulk resolution. A bulk resolution API call is used to resolve taste profile items to pre-stored identifiers associated with a service, such as a service that provides metadata about items associated with the taste profile (e.g., song tempo for a large catalog of items). Yet another example use case for taste profiles is referred to as user-to-user recommendation. This API call is used to discover users with similar tastes by comparing the similarity of taste profile item(s) associated with users.
- A
taste profile 242 can represent a single user or multiple users. Conversely, a single user or entity can have multiple taste profiles 242. For example, one taste profile can be generated in connection with a user's media content play activity, whereas another separate taste profile can be generated for the same user based on the user's selection of media content items and/or artists for a playlist. - Referring still to
FIG. 2 , the mediacontent search server 202 operates to perform media content search in response to a media content search request, such as the user query 120 (FIG. 1 ). In some embodiments, the mediacontent search server 202 includes a mediacontent search application 250, aprocessing device 252, amemory device 254, and anetwork access device 256. Theprocessing device 252, thememory device 254, and thenetwork access device 256 may be similar to theprocessing device 214, thememory device 216, and thenetwork access device 218, respectively, which have each been previously described. - In some embodiments, the media
content search application 250 operates to interact with themedia playback device 102 and provide selection of one or more media content items based on theuser query 120. As described herein, the mediacontent search application 250 can include the descriptive media content search engine 110 (FIG. 1 ). The mediacontent search application 250 can interact with other servers, such as the media content server 200, thevoice interaction server 204, and the user command interpretation server 206, to perform media content search. - Referring still to
FIG. 2 , thevoice interaction server 204 operates to provide various voice-related functionalities to themedia playback device 102. In some embodiments, thevoice interaction server 204 includes avoice recognition application 270, a speech synthesis application 272, aprocessing device 274, amemory device 276, and anetwork access device 278. Theprocessing device 274, thememory device 276, and thenetwork access device 278 may be similar to theprocessing device 214, thememory device 216, and thenetwork access device 218, respectively, which have each been previously described. - In some embodiments, the
voice recognition application 270 and the speech synthesis application 272, either individually or in combination, operate to interact with themedia playback device 102 and enable themedia playback device 102 to perform various voice-related functions, such as voice media content search, voice feedback, voice notifications, etc. - In some embodiments, the
voice recognition application 270 is configured to perform speech-to-text (STT) conversion, such as receiving a recording of voice command (e.g., an utterance) and converting the utterance to a text format. - In some embodiments, the speech synthesis application 272 is configured to perform text-to-speech (TTS) conversion, so that a language text is converted into speech. Then, the
voice interaction server 204 can transmit an audio data or file for the speech to themedia playback device 102 so that themedia playback device 102 generates a voice assistance to the user using the transmitted audio data or file. - Referring still to
FIG. 2 , the user command interpretation server 206 operates to analyze a text version of a user command (e.g., a text version of the utterance) to determine appropriate actions to take according to the user command. - In some embodiments, the user command interpretation server 206 includes a natural language understanding (NLU)
application 280, aprocessing device 282, a memory device 284, and a network access device 286. Theprocessing device 282, the memory device 284, and the network access device 286 may be similar to theprocessing device 214, thememory device 216, and thenetwork access device 218, respectively, which have each been previously described. - In some embodiments, the
NLU application 280 operates to analyze the text format of the utterance to determine functions to perform based on the utterance. TheNLU application 280 can use a natural language understanding algorithm that involves modeling human reading comprehension, such as parsing and translating an input according to natural language principles. -
FIG. 3 is a flowchart of anexample method 300 for selecting and providing media content responsive to auser query 120. It is described herein that themethod 300 is at least partially performed primarily by the media delivery system 104 (e.g., the descriptive mediacontent search engine 110 thereof) that interacts with themedia playback device 102. In other embodiments, themethod 300 can be at least partially performed by themedia playback device 102 that interacts with themedia delivery system 104. In yet other embodiments, themethod 300 can be at least partially performed by themedia playback device 102 and themedia delivery system 104. - At
operation 302, themedia delivery system 104 operates to receive a media content search query, such as the user query 120 (FIG. 1 ). In some embodiments, theuser query 120 can be received using themedia playback device 102. - As described herein, in some embodiments, the
user query 120 is a text typed by the user using themedia playback device 102. In the illustrated example ofFIG. 1 , the user can type in the text (a text form of “play relaxing jazz for tonight”) via themanual input device 160 of themedia playback device 102. In some embodiments, the text of theuser query 120 can be analyzed to identify descriptive terms (e.g., descriptors) that are to be used for media content search as described herein. - In other embodiments, the
user query 120 is an utterance that is received and recorded using themedia playback device 102. In the illustrated example ofFIG. 1 , the user can speak the voice request (an audio form of “play relaxing jazz for tonight”) via thesound detection device 162 of themedia playback device 102. The utterance of theuser query 120 received via themedia playback device 102 can be processed into a text version and further analyzed to identify descriptive terms (e.g., descriptors) that can be used for media content search as described herein. An example method for processing the utterance of theuser query 120 is illustrated and described in more detail herein, for example with reference toFIG. 4 . - At
operation 304, themedia delivery system 104 operates to perform descriptive media content search based on the search query (e.g., the user query 120). In some embodiments, the descriptive media content search can identify media content (e.g., one or more media content items 234) that is most relevant to the descriptive terms (e.g., descriptors) in theuser query 120. In other embodiments, other criteria can be used to determine media content responsive to theuser query 120. An example method for performing such descriptive media content search is illustrated and described in more detail herein, for example with reference toFIG. 5 . - At
operation 306, themedia delivery system 104 operates to provide the identified media content to themedia playback device 102. In some embodiments, themedia delivery system 104 operates to transmit media stream 232 for the identified media content to themedia playback device 102 via thenetwork 106. -
FIG. 4 is a flowchart of anexample method 400 for processing an utterance of theuser query 120. Themethod 400 can also be performed to determine a descriptive media content search service as a fulfillment strategy based on the utterance of theuser query 120. In some embodiments, themethod 400 can be used to at least partially perform theoperation 302 of themethod 300 inFIG. 3 . - In some embodiments, the
method 400 can be performed by themedia delivery system 104. For example, themethod 400 can be used by the natural language understanding (NLU) system (e.g., the NLU application 280) for performing a fulfillment strategy based on a received utterance. As described herein, themedia delivery system 104 can include the user command interpretation server 206 that operates theNLU application 280, and themedia playback device 102 can receive an utterance of theuser query 120 and provide the utterance to themedia delivery system 104 for processing at the user command interpretation server 206. This is for example purposes only, and other configurations are possible. For instance, theNLU application 280 is locally saved and performed in themedia playback device 102. - The
method 400 can begin at operation 410 in which themedia delivery system 104 receives the utterance of theuser query 120. The utterance of theuser query 120 can be stored and received asutterance data 412 by themedia delivery system 104. Theutterance data 412 is data describing the utterance of theuser query 120. In at least some embodiments, theutterance data 412 is an audio recording that contains the utterance being spoken. In some examples, theutterance data 412 is received as an entire audio data file. For instance, themedia playback device 102 buffers theutterance data 412 as it is obtained from thesound detection device 162. The bufferedutterance data 412 is then sent to themedia delivery system 104 for processing. In other instances, themedia playback device 102 streams theutterance data 412 to themedia delivery system 104 in real-time as theutterance data 412 is received from thesound detection device 162 of themedia playback device 102. In some examples, theutterance data 412 is stored (e.g., by the media delivery system 104) in a data store after it is received. After theutterance data 412 is received, themethod 400 moves tooperation 420. -
Operation 420 includes performing automated speech recognition on theutterance data 412 to obtaintext data 422. In some embodiments, performing automated speech recognition includes providing theutterance data 412 as input to an automated speech recognition system (e.g., thevoice recognition application 270 of the voice interaction server 204) and receiving thetext data 422 as output from the automated speech recognition system. Automated speech recognition can be performed using any of a variety of techniques (e.g., using hidden Markov models or neural networks). Examples of automated speech recognition systems include CMU SPHINX, maintained by CARNEGIE MELLON UNIVERSITY, and DEEPSPEECH, maintained by the MOZILLA FOUNDATION. After thetext data 422 is obtained from the automated speech recognition system, themethod 400 can move tooperation 430. -
Operation 430 includes determining aslot 434 and an intent 432 from thetext data 422. Theslot 434 is a key-value pair that describes a portion of thetext data 422 having a specific meaning. The intent 432 describes a general intent of thetext data 422. In the illustrated example ofFIG. 1 , if thetext data 422 represents “play relaxing jazz for tonight” as input, the intent 432 is “play” and theslot 434 can include at least one key-value pair, such as {descriptor:relaxing} and {descriptor:tonight}. In another example, if thetext data 422 includes “play Thriller” as input, the intent 432 is “play” and theslot 434 is a key-value pair of {song: Thriller}. - In some embodiments, the
text data 422 can include asingle slot 434 and asingle intent 432. In other embodiments, theoperation 430 can return a plurality ofslots 434 and/or a plurality ofintents 432 from thetext data 422. In yet other embodiments, thetext data 422 provides an intent 432 but noslot 434. For example, where thetext data 422 is “play,” theoperation 430 can return the intent 432 being “play”, but will not result in any slot 434 (e.g., thetext data 422 does not include a description of what to play). In other examples, the text data 433 can include one ormore slots 434 but no intent. For instance, where thetext data 422 is “All Along the Watchtower by Jimi Hendrix,” theoperation 430 can result in two slots 434 (e.g., {Song: All Along the Watchtower} and {Artist: Jimi Hendrix}) but no intent 432 (e.g., thetext data 422 does not include a description of what do to with the song and artist, such as search, play, or save). - As described herein, the
text data 422 can include one or more descriptors. In some embodiments, the descriptors can include values of one or more of theslots 434 identified from thetext data 422. In the illustrated example ofFIG. 1 , where thetext data 422 includes “play relaxing jazz for tonight” converted from the utterance of theuser query 120, the values of theslots 434, such as “relaxing,” “relaxing jazz,” and/or “tonight,” can be included as descriptors of thetext data 422. In other embodiments, the descriptors of thetext data 422 can include the value of the intent 432 (e.g., “play” in the above example). In yet other embodiments, other terms identified in the text data 422 (e.g., “jazz”) can be included in the descriptors of thetext data 422. In yet other embodiments, the descriptors of thetext data 422 can be identified in other methods. - In some embodiments, the
operation 430 can be performed by a natural language understanding model that is trained to identify theslot 434 andintent 432 for thetext data 422 provided as input. The natural language understanding model can be implemented in a variety of ways, including using a state vector machine or a conditional random fields model, among others. With the intent 432 and theslots 434 determined, themethod 400 can move tooperation 440. -
Operation 440 includes determining afulfillment strategy 442 using theslot 434 and the intent 432. Thefulfillment strategy 442 is a course of action to be performed to execute a command or service associated with the utterance, such as theintent 432 thereof. As described herein, thefulfillment strategy 442 can include a descriptive search service in response to theuser query 120. In another example, where the intent 432 is a play intent, thefulfillment strategy 442 is a play fulfillment strategy and involves execution of a play command. - In some embodiments, the
media delivery system 104 includes a fulfillment manager that operates to select afulfillment strategy 442 among a plurality of fulfillment strategies, which satisfies the received utterance. The plurality of fulfillment strategies can include a play fulfillment strategy, a basic search strategy (e.g., using entity-focused search), a descriptive search strategy, a recommendation strategy, and an unspecified search strategy. In some examples, the fulfillment manager follows a decision tree based the intent 432 and theslot 434. In other examples, eachfulfillment strategy 442 defines requirements (e.g., a play fulfillment strategy may require a play intent) and the fulfillment manager selects a fulfillment strategy among the plurality of fulfillment strategies based on whether the associated requirement is met. - Once the
fulfillment strategy 442 is selected, thefulfillment strategy 442 is then performed by themedia delivery system 104. For example, where thefulfillment strategy 442 is determined to be a descriptive search service based on the intent 432 and/or theslot 434 of thetext data 422, themedia delivery system 104 performs descriptive media content search based on the text data 422 (e.g., at least one of the descriptors identified from the text data 422). -
FIG. 5 is a flowchart of anexample method 500 for performing descriptive search for media content based on theuser query 120. In some embodiments, themethod 500 can be used to at least partially perform theoperation 304 of themethod 300 as illustrated inFIG. 3 . - At
operation 502, themedia delivery system 104 can operate to identify descriptive terms associated with candidate media content items. As described herein, the candidate media content items can include a subset of media content items from media content item database and are subject to the search in response to theuser query 120. In some embodiments, the candidate media content items are obtained from candidate playlists 558 (FIG. 7 ). - In some embodiments, the
operation 502 is performed to set up a descriptive search database that is to be utilized to conduct a descriptive search. In some embodiments, theoperation 502 can be performed to build and store the descriptive search database prior to receiving theuser query 120. In other embodiments, theoperation 502 is performed upon receiving theuser query 120. An example of theoperation 502 is illustrated and described in more detail herein, for example with reference toFIG. 6 . - At
operation 504, themedia delivery system 104 can operate to select one or more media content items from the candidate media content items. In some embodiments, theoperation 504 is performed to utilize the descriptive search database and conduct a descriptive search based on one or more descriptors of theuser query 120. An example of theoperation 504 is illustrated and described in more detail herein, for example with reference toFIG. 12 . -
FIG. 6 is a flowchart of anexample method 530 for generating or updating descriptive search database. In some embodiments, themethod 530 is used to at least partially perform theoperation 502 as illustrated inFIG. 5 . As described herein, the descriptive search database can be generated by identifying descriptive terms associated with media content items. - In general, the descriptive search solution according to an exemplary embodiment of the present disclosure utilizes a corpus of playlists, each of which contains a playlist description that describe the respective playlist. The playlist description of a playlist can be used to identify additional descriptive information about media content items in the playlist, and use that information to provide relevant search results.
- At
operation 532, themedia delivery system 104 operates to retrieve candidate playlists 558 (FIG. 7 ). Thecandidate playlists 558 includes a subset of playlists selected from playlist database and subject to the search as described herein. An example of theoperation 532 is illustrated and described in more detail herein, for example with reference toFIG. 7 . - At
operation 534, themedia delivery system 104 operates to retrieve adescriptive term lexicon 560 as illustrated inFIG. 8 . Thedescriptive term lexicon 560 includes a list of words or phrases that is prepopulated and used as a vocabulary to identify descriptive terms fromplaylist descriptions 570 in the candidate playlists 558 (including 558A, 558B, and 558C), as illustrated inFIG. 8 . Thedescriptive term lexicon 560 can be used to clarify unclear terms in the playlist descriptions of thecandidate playlists 558 and identify them as descriptive terms relevant to the descriptive search. In some embodiments, thedescriptive term lexicon 560 can be manually created and updated. In other embodiments, thedescriptive term lexicon 560 can be at least partially automatically created and updated using, for example, machine learning technology. An example of theoperation 534 is illustrated and described in more detail herein, for example with reference toFIG. 8 . Theoperation 534 can be optional and, in other embodiments, can be omitted from themethod 530. - At
operation 536, themedia delivery system 104 operates to analyze aplaylist description 570 of eachcandidate playlist 558 and identify one or more descriptive terms 572 (including 572A, 572B, and 572C) from theplaylist description 570. In some embodiments, theplaylist description 570 is analyzed using thedescriptive term lexicon 560. In some embodiments, theplaylist description 570 is tokenized into tokenized terms (e.g., tokens), and at least one of the tokenized terms can be determined as the descriptive terms 572. An example of theoperation 536 is illustrated and described in more detail herein, for example with reference toFIG. 8 . - At
operation 538, themedia delivery system 104 operates to associate the descriptive terms 572 identified from eachcandidate playlist 558 with each of themedia content items 234 in thecandidate playlist 558. Themedia content items 234 associated with thecandidate playlist 558 can be identified by media content item identifications (IDs) 578. - As such, the descriptive terms 572 associated with the media content items are used as additional information descriptive of the media content items. An example association of the descriptive terms 572 with the media content items is illustrated and described in more detail herein, for example with reference to
FIG. 9 . - At
operation 540, themedia delivery system 104 operates to weight themedia content items 234 associated with the descriptive terms 572. In some embodiments, eachmedia content item 234 associated with a given descriptive term 572 can be weighted to provide an indication of how strongly themedia content item 234 is associated with the descriptive term 572. In other embodiments, the weighting can also be based at least in part on how frequency themedia content item 234 appears in the candidate playlists that contain the descriptive term 572 in theplaylist description 570. In one example, a term frequency and inverse document frequency (TFIDF) weighting is used to compute a weighting score for each of themedia content items 234 from thecandidate playlists 558. An example of theoperation 540 is illustrated and described in more detail herein, for example with reference toFIG. 10 . Theoperation 540 can be optional and, in other embodiments, can be omitted from themethod 530. - At
operation 542, themedia delivery system 104 operates to generate and/or update a descriptive search database 610 (FIG. 11 ). When the descriptive terms 572 are identified, the descriptive terms 572 can be stored in thedescriptive search database 610. In some embodiments, an inverted index is created to map the descriptive terms 572 to themedia content items 234 in thecandidate playlists 558. For example, thedescriptive search database 610 stores each of the descriptive terms 572 and includes a list of media content items associated with that descriptive term 572. An example data structure of thedescriptive search database 610 is illustrated and described in more detail herein, for example with reference toFIG. 11 . -
FIG. 7 illustrates anexample method 548 for determiningcandidate playlists 558. In some embodiments, themethod 548 is used to at least partially perform theoperation 532 as described inFIG. 6 . - In general, the
method 548 can be performed to identify a subset of playlists from aplaylist database 552. Theplaylist database 552 includes a plurality ofplaylists 554, which can include the playlists 239 as described inFIG. 2 . The subset of playlists can be referred to herein ascandidate playlists 558. In some embodiments, acandidate playlist database 556 is created to store thecandidate playlists 558. Themethod 548 is designed to filter out some of theplaylists 554 in theplaylist database 552 that have less coherent or suitable to be used for descriptive search. By way of example, some of theplaylists 554 in theplaylist database 552 can have no or little playlist description, or playlist descriptions with no meaningful information that characterize the playlists. Theseplaylists 554 in theplaylist database 552 are not useful for performing themethod 530 as described herein. Themethod 548 is performed to selectcandidate playlists 558 that are meaningful, suitable playlists that can lead to relevant search results in response to theuser query 120. - In some embodiments, the
method 548 can be performed at least by a candidateplaylist selection engine 550 that can be included in themedia delivery system 104. The candidateplaylist selection engine 550 can operate to retrieve theplaylists 554 stored in theplaylist database 552. In other embodiments, only some of theplaylists 554 in theplaylist database 552 are retrieved. - The candidate
playlist selection engine 550 can use one or more various algorithms to selectcandidate playlists 558. In some embodiments, a heuristic technique can be used to selectcandidate playlists 558 from theplaylists 554. In other embodiments, thecandidate playlists 558 can be selected based on a predetermined rule. - The
candidate playlists 558 in thecandidate playlist database 556 can be updated as theplaylists 554 in theplaylist database 552 change. For example, theplaylists 554 can change in theplaylist database 552 as users create new playlists, edit or remove existing playlists, and add, remove, and modify media content items in existing playlists. In some embodiments, thecandidate playlists 558 can be updated periodically, in real-time, and/or at preset times. -
FIG. 8 illustrates anexample method 562 for identifying one or more descriptive terms 572 from thecandidate playlists 558. In some embodiments, themethod 562 is used to at least partially perform theoperations FIG. 6 . - As described herein, a
candidate playlist 558 has aplaylist description 570. In some embodiments, theplaylist description 570 includes aplaylist title 574 and adescriptive text 576. As described herein, theplaylist title 574 is a name of the playlist, which can be created either by a user or by a media streaming service provider (either manually or automatically). Thedescriptive text 576 can include words, phrases, or sentences that characterize thecandidate playlist 558. Thedescriptive text 576 can be obtained from various sources. In some examples, a user can provide thedescriptive text 576. In other examples, thedescriptive text 576 can be obtained from external sources, such as user or expert opinions or feedback (e.g., reviews or genre labels through websites, chatrooms, blogs, surveys, and the like), statistics (e.g., sales data), or any text information that may be used to describe, rank, or interpret the playlist or music associated with the playlist. In some embodiments, thedescriptive text 576 can also be generated by a community of listeners and automatically retrieved from Internet sites, chat rooms, blogs, and the like. - For each
candidate playlist 558, theplaylist description 570 can be analyzed to identify one or more descriptive terms 572. In some embodiments, theplaylist description 570 is tokenized into tokenized terms (e.g., tokens), at least some of which are determined as descriptive terms 572. For example, for a givencandidate playlist 558, theplaylist description 570 is analyzed such that one or more keywords in theplaylist description 570 are identified and stored as tokens. In some embodiments, thedescriptive term lexicon 560 is used in identifying keywords from theplaylist description 570 and determining the keywords as the descriptive terms 572. - By way of example, in the illustrated example of
FIG. 8 , afirst candidate playlist 558A contains aplaylist description 570 including a playlist title 574 (“Relaxing Night”) and a descriptive text 576 (“Great for relaxation afterwork”). Themedia delivery system 104 can process (e.g., tokenize) theplaylist description 570 to identify keywords, such as “relaxing,” “night,” “great,” “relaxation,” and “afterwork” and generate tokens for such keywords and/or their variants. At least one of the tokens can be included in thedescriptive terms 572A for thefirst candidate playlist 558A. Second andthird candidate playlists descriptive terms - The
playlist description 570 can be tokenized in various ways. In some embodiments, theplaylist description 570 can be tokenized by a single word. In other embodiments, theplaylist description 570 can be tokenized by a combination of adjacent words if such adjacent words are determined to be related. Other methods are also possible. -
FIG. 9 illustrates anexample data structure 580 that associates the descriptive terms 572 withmedia content items 234. In some embodiments, thedata structure 580 is generated from theoperation 538 as illustrated inFIG. 6 . - In some embodiments, the
data structure 580 is configured as a data table having anitem title column 582, anitem ID column 584, aplaylist ID column 586, adescriptive term column 588, and one ormore metadata columns 590. Theitem title column 582 shows the titles of themedia content items 234. Theitem ID column 584 shows the ID (e.g., a uniform resource identifier (URI)) of themedia content items 234. Theplaylist ID column 586 shows the IDs of the playlists that contain respectivemedia content items 234. Thedescriptive term column 588 shows the descriptive terms 572 associated with respectivemedia content items 234. As described herein, the descriptive terms 572 for a givenmedia content item 234 can be identified from the playlists (e.g., the candidate playlists 558) containing thatmedia content item 234. - As illustrated, each of the
media content items 234 from thecandidate playlists 558 are associated with the descriptive terms 572 identified from thecandidate playlists 558. By way of example, amedia content item 234 titled “ABC” is identified as item ID “123456abw” and can be associated with the descriptive terms, “relax” or its variants (e.g., “relaxing” or “relaxation”), “great,” “night” or its variants (e.g., “tonight”), “afterwork,” “kid,” “favorite,” and “good,” which are identified from thecandidate playlists media content item 234 is associated with the descriptive term “relax” or its variants twice from thecandidate playlists candidate playlists - The association of the descriptive terms 572 identified from the
candidate playlists 558 with themedia content items 234 in thecandidate playlists 558 provides additional descriptive information to themedia content items 234. The idea of this association is that the descriptive terms of a playlist is an indication of the characteristics of the media content items contained in that playlist. Therefore, a playlist description of a playlist can suggest the characteristics of the media content items in the playlist. - In some embodiments, the
data structure 580 also includes at least part of the information from metadata (e.g., the metadata 236 (FIG. 2 )) of themedia content items 234. Such metadata information is shown in one or more of themetadata columns 590. In the illustrated example ofFIG. 9 , themetadata columns 590 includes a column for genre of themedia content items 234. -
FIG. 10 is a flowchart of anexample method 600 for weighting themedia content items 234 associated with the descriptive terms 572. In some embodiments, themethod 600 is used to at least partially perform theoperation 540 as described inFIG. 6 . In some embodiments, themethod 600 can be performed based on the data that map themedia content items 234 to the descriptive terms 572, such as shown in thedata structure 580 inFIG. 9 . - In some embodiments, the
method 600 can be used to compute how relevant a givenmedia content item 234 is to theuser query 120. Themethod 600 can provide a solution to ranking themedia content items 234 in view of theuser query 120. In some embodiments, themethod 600 can determine how relevant or important a given descriptive term is to a givenmedia content item 234 among a plurality ofmedia content items 234. - The
method 600 can begin atoperation 602 in which themedia delivery system 104 operates to calculate afirst term frequency 640. Thefirst term frequency 640 is a term frequency in each of themedia content items 234. Thefirst term frequency 640 is a measure of how frequently a particular term (e.g., a descriptive term) appears in a document (e.g., a media content item). In some embodiments, the first term frequency (TF) 640 can be computed as: -
- At
operation 604, themedia delivery system 104 operates to calculate asecond term frequency 642. Thesecond term frequency 642 is a measure of how frequently a particular term (e.g., a descriptive term) appears in all documents (e.g., all candidate media content items). In some embodiments, the second term frequency (IDF) 642 can be computed as: -
- It is noted that IDF=0 if the term does not appear in any media content item.
- At
operation 606, themedia delivery system 104 operates to calculate arelevance score 644 based on thefirst term frequency 640 and thesecond term frequency 642. Therelevance score 644 is a measure of how relevant a particular term (e.g., a descriptive term) is to a document (e.g., a media content item). In some embodiments, therelevance score 644 can be computed as: -
Relevance Score (descriptive term, media content item)=TF (descriptive term, media content item)×IDF (descriptive term) - In some embodiments, the
method 600 implements a term frequency and inverse document frequency (TFIDF) weighting. In other embodiments, other weighting algorithms can be used for themethod 600. -
FIG. 11 illustrates an example of thedescriptive search database 610. In some embodiments, thedescriptive search database 610 is generated and/or updated at least partially by theoperation 542 as described inFIG. 6 . - In some embodiments, the
descriptive search database 610 includes aninverted index structure 620 that maps the descriptive terms 572 to themedia content items 234 that are contained in thecandidate playlists 558. Other indexing methods are also possible in thedescriptive search database 610. - In some embodiments, the
inverted index structure 620 is configured as a data table having adescriptive term column 622, anitem title column 624, anitem ID column 626, arelevance score column 628, and one ormore metadata columns 630. Thedescriptive term column 622 includes the descriptive terms 572 identified from thecandidate playlists 558. Theitem title column 624 shows the titles of themedia content items 234. Theitem ID column 626 shows the ID (e.g., a uniform resource identifier (URI)) of themedia content items 234. Therelevance score column 628 includes relevance scores calculated for each descriptive term 572 with respect to eachmedia content item 234. In some embodiments, the relevance scores can be obtained by themethod 600 as described inFIG. 10 . Themetadata columns 630 includes at least part of the information from metadata (e.g., the metadata 236 (FIG. 2 )) of themedia content items 234. As described herein, the metadata of themedia content items 234 can be additionally used to improve the search results. -
FIG. 12 is a flowchart of anexample method 700 for performing a descriptive search from thedescriptive search database 610. In some embodiments, themethod 700 is used to at least partially perform theoperation 504 as illustrated inFIG. 5 . - In some embodiments, the
method 700 is used to select one or more media content items from the candidate media content items (e.g., the media content items from the candidate playlists 558) by utilizing the descriptive search database and conducting a descriptive search based on one or more descriptors of theuser query 120. - The
method 700 can begin atoperation 702 in which themedia delivery system 104 operates to process theuser query 120 to identify one or more search keywords 734 (FIG. 13 ). In some embodiments, theuser query 120 is analyzed to identify one ormore descriptors 736, and one or more of thedescriptors 736 from theuser query 120 can be used for one ormore search keywords 734. An example of theoperation 702 is illustrated and described in more detail herein, for example with reference toFIG. 13 . - At
operation 704, themedia delivery system 104 retrieves thedescriptive search database 610. Atoperation 706, themedia delivery system 104 operates to select one or more media content items 740 (FIG. 14 ) (also referred to herein as output media content items or the like) based on therelevance score 644 associated with thesearch keywords 734. An example of theoperations FIG. 14 . - At
operation 708, themedia delivery system 104 retrieves user profile data 776 (FIG. 15 ). Atoperation 710, themedia delivery system 104 operates to select one or more personalizedmedia content items 772 based on theuser profile data 776. An example of theoperations FIG. 15 . - In some embodiments, at least one of the
operations operation 502 as illustrated inFIG. 5 while the rest of theoperations operation 502. In other embodiments, theoperations operation 502 as illustrated inFIG. 5 . -
FIG. 13 illustrates anexample method 730 for analyzing theuser query 120 to identify one ormore search keywords 734. In some embodiments, themethod 730 is used to at least partially perform theoperation 702 as described inFIG. 12 . As described herein, theuser query 120 is a text query that is typed by the user, or a voice request (e.g., an utterance) received through a microphone and converted into a text query. - In some embodiments, the
method 730 can be performed at least by a searchquery process engine 732 that can be included in themedia delivery system 104. The searchquery process engine 732 can operate to analyze theuser query 120 and generate thesearch keywords 734. In some embodiments, theuser query 120 is tokenized into tokenized terms (e.g., words or phrases), which includedescriptors 736. At least one of thedescriptors 736 can be used as thesearch keywords 734. - By way of example, the
user query 120 is a search request “play relaxing jazz for tonight.” The searchquery process engine 732 operates to process theuser query 120 and identify thesearch keywords 734, such as “relaxing” (or its variants), “tonight” (or its variants), “jazz” (or its variants), etc. - In some embodiments, the
search keywords 734 include at least one of thedescriptors 736 from theuser query 120, such asdescriptive search descriptors 738 andother descriptors 739. In some embodiments, thedescriptive search descriptors 738 are primarily used to perform a descriptive search as described herein. Theother descriptors 739 can be used to perform other types of search, such as entity-focused search using media content metadata. As described herein, theother descriptors 739 can be used in addition to thedescriptive search descriptors 738 to improve the search results from the descriptive search responsive to theuser query 120. - In some embodiments, the search
query process engine 732 refers to thedescriptive search database 610 to identify thesearch keywords 734 by checking to see if any of the tokenized terms (or the descriptors) in theuser query 120 match any of the descriptive terms in thedescriptive search database 610. If so, such tokenized terms from theuser query 120 are identified as search keywords to be used in the descriptive search. -
FIG. 14 illustrates anexample method 750 for selectingmedia content items 740 based on descriptive search utilizing thedescriptive search database 610. In some embodiments, themethod 750 is used to at least partially perform theoperations FIG. 12 . - In some embodiments, the
method 750 can be performed at least by adescriptive search engine 752 that can be included in themedia delivery system 104. Thedescriptive search engine 752 can process thesearch keywords 734 based on thedescriptive search database 610 and return one or moremedia content items 740 that match theuser query 120. - In some embodiments, the
descriptive search engine 752 selects themedia content items 740 based on therelevance score 644. For example, themedia content items 740 are selected to match thesearch keywords 734 and have the highest relevance scores 644. In some embodiments, a predetermined number ofmedia content items 740 are selected having the highest relevance scores 644 (e.g., top 10 media content items). In other embodiments, the number ofmedia content items 740 are determined based on other factors, such as a relevance score threshold. - In other embodiments, the
descriptive search engine 752 can employ other algorithms to select themedia content items 740 in response to theuser query 120. For example, thedescriptive search engine 752 can operate to obtain a relevance score vector for each of themedia content items 234 from thecandidate playlists 558. Further, thedescriptive search engine 752 can operate to obtain a corresponding vector for theuser query 120. Then, thedescriptive search engine 752 operates to compare each of the relevance score vectors for themedia content items 234 with the corresponding vector for theuser query 120, and determine similarity based on the comparison. In some embodiments, cosine similarity can be used to determine the similarity between the vectors. Thedescriptive search engine 752 can selectmedia content items 740 having relevance score vectors with the highest similarities with the corresponding vector for theuser query 120. Other algorithms are also possible in other embodiments. - In some embodiments, where a plurality of
search keywords 734 are used for search, therelevance score 644 for a media content item containing the plurality ofsearch keywords 734 can be computed by multiplication of individual relevance scores 644 for each of thesearch keywords 734 associated with the media content item. An example of this case is illustrated with the table 620 inFIG. 14 . - In some embodiments, the
descriptive search engine 752 can utilize other information about themedia content items 234, such as information from metadata (e.g., the metadata 236) of themedia content items 234, to provide even more relevant search results. In the illustrated example ofFIG. 14 , genre information of themedia content items 234 are used in view of the other descriptors 739 (e.g., “jazz”) in thesearch keywords 734. For example, in addition to the descriptive search described herein, thedescriptive search engine 752 can further operate to search formedia content items 234 having information from the metadata that matches theother descriptors 739 of thesearch keywords 734. As such, the metadata ofmedia content items 234 can improve the search results from the descriptive search. By way of another example, where a search query includes a combination of descriptive search terms and other typical descriptors, such as “play some relaxing jazz music by Miles Davis,” in which case the typical music search algorithm (e.g., using media content item metadata) can be used to identify music by Miles Davis, and the descriptive search algorithm can be used to identify “relaxing” music by the same artist. -
FIG. 15 illustrates anexample method 770 for selecting personalizedmedia content items 772 from themedia content items 740. In some embodiments, themethod 770 is used to at least partially perform theoperations FIG. 12 . - In some embodiments, the
method 770 can be performed at least by apersonalization engine 774 that can be included in themedia delivery system 104. Thepersonalization engine 774 can select media content items that are personalized to the user based at least in part onuser profile data 776. As such, themethod 770 enables the search results to be personalized based on theuser profile data 776. - In some embodiments, the
personalization engine 774 utilizes theuser profile data 776 to select one or more media content items from themedia content items 740 that are selected by the descriptive search as described herein. Thepersonalization engine 774 then outputs the selected media content items as personalizedmedia content items 772. - In some embodiments, the
user profile data 776 include user information. The user information can include one or more user specific features that can be used to determine media content personalized to respective users. The user specific features can represent the user's preference, media content consumption history, and/or other general features associated with the user of themedia playback device 102. A few examples of user specific features include various user media content consumption activities (e.g., listening activities), such as the most listened context (e.g., track, album, artist, playlist, etc.), the most preferred context (e.g., track, album, artist, playlist, etc.), and the most preferred genre or other acoustic feature. The user specific features can be obtained based on the user's long-term activities (e.g., over a few months, a year or longer, etc.) or short-term activities (e.g., over a year or less, a month, a week, etc.). - In some embodiments, the user information includes user biographical information and user taste profile information. The user biographical information include features relating to personal information of respective users, such as address, age, gender, birth date/year, family, relationships, profession, education, religious preference, sexual preference, association, and other information associated with users. In some embodiments, the user biographical information can be obtained at least partially from the user account information.
- The user taste profile information include features that can be obtained from
taste profiles 242 of respective users, which represent the user's activities with respect to media content consumption, such as user preferences and history of consuming media content, as described herein. As described herein, the user taste profile information includes a user listening history, and other user-related data that can be used to personalize the search results. In some embodiments, the user taste profile information can further include taste profiles of other users having similar tastes. - As illustrated in
FIG. 15 as an example, theuser profile data 776 can provide scores of one or more user activities with respect to each of the media content items 740 (e.g., “ABC” and “BCD”). By way of example, theuser profile data 776 for the particular user provides a playback history score for eachmedia content item 740 to represent how often the user has played themedia content item 740. Theuser profile data 776 can also provide an artist preference score to represent how much the user likes the artist of each of themedia content items 740. In some embodiments, where a plurality of scores are considered, theuser profile data 776 can provide a single score (“Composite Score”) that represents all the scores. In some embodiments, such a single score can be calculated by multiplication of the plurality of scores. - Although the systems and the methods according to the present disclosure are primarily described to return
media content items - In certain examples, an exemplary embodiment of the present disclosure operate to receive an utterance from a voice-enabled device. The embodiment can analyze a text version of the utterance to determine one or more descriptors in the utterance. The embodiment can optionally sort out a subset of playlists from a corpus of playlists in database based on one or more criteria (e.g., playlist size, time since last modified, various patterns to the playlist title, and other criteria). For each of the subset of playlists, the embodiment can tokenize the playlist title and/or other description into terms. For each of the subset of playlists, the embodiment can tag the media content items in the playlist with the tokenized terms so that the media content items are regarded as keys for the tokenized terms. The embodiment can aggregate the tags in each of the media content items in the entire subset of playlists. For example, the embodiment can group by the media content items to collect all the tokenized terms (including duplicates) associated with each of the media content items. The embodiment can apply term frequency and inverse document frequency (TFIDF) to weight the tokenized terms among the different media content items. The embodiment can generate an inverted index that maps the tokenized terms to the media content items. The embodiment can perform search for media content items using the inverted index based on the descriptors in the utterance. For example, the search can be performed by comparing the descriptors in the utterance with the tokenized terms in the inverted index, and then determining one or more of the tokenized terms in the inverted index and identifying the media content items associated with the determined tokenized terms. The embodiment can then return the identified media content items to the voice-enabled device for playback or other controlling of playback.
- The various examples and teachings described above are provided by way of illustration only and should not be construed to limit the scope of the present disclosure. Those skilled in the art will readily recognize various modifications and changes that may be made without following the examples and applications illustrated and described herein, and without departing from the true spirit and scope of the present disclosure.
Claims (21)
1-20. (canceled)
21. A method for providing media content, the method comprising:
receiving, at a media delivery system having one or more servers, a media content search query with a keyword from a media playback device;
processing, by the media delivery system, the media content search query to identify the keyword;
selecting, by the media delivery system, media content based on a relevance score associated with the keyword and assigned to the media content, the relevance score being based at least in part on a frequency of the keyword in playlist descriptions of playlists having the media content; and
providing a media stream for the media content to the media playback device.
22. The method of claim 21 , wherein the relevance score is calculated at least in part by:
calculating a first term frequency, the first term frequency associated with a first frequency of the keyword appearing in the playlist descriptions of the playlists having the media content;
calculating a second term frequency, the second term frequency associated with a number of playlists containing the keyword in the playlist description; and
calculating the relevance score for the media content based on the first term frequency and the second term frequency.
23. The method of claim 21 , wherein the relevance score is calculated at least part by using term frequency and inverse document frequency (TFIDF).
24. The method of claim 21 , wherein the playlist descriptions include a playlist title and a descriptive text.
25. The method of claim 24 , wherein the descriptive text is obtained from at least one of user reviews, expert reviews, and statistics.
26. The method of claim 21 , further comprising:
retrieving a user profile associated with a user who provided the media content search query,
wherein selecting the media content is based at least in part on the user profile.
27. The method of claim 21 , further comprising:
retrieving a descriptive term lexicon, wherein the playlist descriptions are tokenized into one or more keywords based on the descriptive term lexicon.
28. The method of claim 21 , further comprising:
retrieving metadata of the media content, wherein the media content is selected based in part on the metadata of the media content.
29. A media delivery system comprising:
at least one processing device; and
at least one data storage device storing data instructions executable by the at least one processing device to cause the media delivery system to:
receive a media content search query with a keyword from a media playback device;
process the media content search query to identify the keyword;
select, media content based on a relevance score associated with the keyword and assigned to the media content, the relevance score being based at least in part on a frequency of the keyword in playlist descriptions of playlists having the media content; and
provide a media stream for the media content to the media playback device.
30. The media delivery system of claim 29 , wherein the instructions further cause the media delivery system to:
calculate a first term frequency, the first term frequency associated with a first frequency of the keyword appearing in the playlist descriptions of the playlists having the media content; and
calculate a second term frequency, the second term frequency associated with a number of playlists containing the keyword in the playlist description,
wherein the relevance score for the media content is calculated at least in part based the first term frequency and the second term frequency.
31. The media delivery system of claim 29 , wherein the relevance score for the media content is calculated using term frequency and inverse document frequency (TFIDF).
32. The media delivery system of claim 29 , wherein the media content search query includes voice data and instructions further cause the media delivery system to:
convert the voice data into text data; and
identify the keyword from the text data.
33. The media delivery system of claim 32 , wherein to identify the keyword from the text data includes to:
determine an intent and at least one slot from the text data; and
identify a descriptor including the keyword from at least one of the intent and the at least one slot.
34. A media playback device comprising:
at least one processor;
at least one memory storing instructions which, when executed by the at least one processor, cause the media playback device to:
receive a receive a media content search query with a keyword;
provide the media content search query to a media delivery system having one or more servers; and
receive media content from the media delivery system in response to the media content search query,
wherein the media delivery system is configured to select the media content based on a relevance score associated with the keyword and assigned to the media content, the relevance score being based at least in part on a frequency of the keyword in playlist descriptions of playlists having the media content.
35. The media playback device of claim 34 , wherein the media content search query is a voice request from a user.
36. The media playback device of claim 35 , wherein the instructions further cause the media playback device to:
convert the voice request into text data; and
generate the media content search query based on the text data.
37. The media playback device of claim 35 , wherein to provide the media content search query to the media delivery system comprises providing the voice request to the media delivery system for processing.
38. The media playback device of claim 34 , wherein the media playback device provides a voice assistant service.
39. The media playback device of claim 34 , wherein the media content includes a playlist of media content items.
40. The media playback device of claim 34 , wherein the media content is personalized based at least in part on user profile data of a user account that is logged in at the media playback device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US17/993,720 US20230205809A1 (en) | 2018-05-24 | 2022-11-23 | Descriptive media content search |
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/988,943 US11163817B2 (en) | 2018-05-24 | 2018-05-24 | Descriptive media content search |
US17/489,937 US11537651B2 (en) | 2018-05-24 | 2021-09-30 | Descriptive media content search |
US17/993,720 US20230205809A1 (en) | 2018-05-24 | 2022-11-23 | Descriptive media content search |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/489,937 Continuation US11537651B2 (en) | 2018-05-24 | 2021-09-30 | Descriptive media content search |
Publications (1)
Publication Number | Publication Date |
---|---|
US20230205809A1 true US20230205809A1 (en) | 2023-06-29 |
Family
ID=68613684
Family Applications (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/988,943 Active 2038-10-15 US11163817B2 (en) | 2018-05-24 | 2018-05-24 | Descriptive media content search |
US17/489,937 Active US11537651B2 (en) | 2018-05-24 | 2021-09-30 | Descriptive media content search |
US17/993,720 Pending US20230205809A1 (en) | 2018-05-24 | 2022-11-23 | Descriptive media content search |
Family Applications Before (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/988,943 Active 2038-10-15 US11163817B2 (en) | 2018-05-24 | 2018-05-24 | Descriptive media content search |
US17/489,937 Active US11537651B2 (en) | 2018-05-24 | 2021-09-30 | Descriptive media content search |
Country Status (1)
Country | Link |
---|---|
US (3) | US11163817B2 (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10645455B1 (en) | 2017-07-17 | 2020-05-05 | Pandora Media, Llc | Delivering artist messages to listeners based on predicted responses |
US11163817B2 (en) | 2018-05-24 | 2021-11-02 | Spotify Ab | Descriptive media content search |
US11106728B2 (en) * | 2019-05-14 | 2021-08-31 | Accenture Global Solutions Limited | Artificial intelligence based music playlist reordering and song performance assessment |
US11314827B2 (en) * | 2019-08-28 | 2022-04-26 | Houzz, Inc. | Description set based searching |
CN111611492A (en) * | 2020-05-26 | 2020-09-01 | 北京字节跳动网络技术有限公司 | Method and device for triggering search, electronic equipment and storage medium |
US20220116399A1 (en) * | 2020-10-12 | 2022-04-14 | Apple Inc. | Media service configuration |
US20220382766A1 (en) * | 2021-06-01 | 2022-12-01 | Apple Inc. | Automatic Media Asset Suggestions for Presentations of Selected User Media Items |
CN113889146A (en) * | 2021-09-22 | 2022-01-04 | 北京小米移动软件有限公司 | Audio recognition method and device, electronic equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040117189A1 (en) * | 1999-11-12 | 2004-06-17 | Bennett Ian M. | Query engine for processing voice based queries including semantic decoding |
US20050071328A1 (en) * | 2003-09-30 | 2005-03-31 | Lawrence Stephen R. | Personalization of web search |
US20090094231A1 (en) * | 2007-10-05 | 2009-04-09 | Fujitsu Limited | Selecting Tags For A Document By Analyzing Paragraphs Of The Document |
US20120303713A1 (en) * | 2001-04-19 | 2012-11-29 | Harbick Andrew V | Playlist-based detection of similar digital works and work creators |
US20160360266A1 (en) * | 2015-06-05 | 2016-12-08 | Google Inc. | Video channel categorization schema |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7546288B2 (en) | 2003-09-04 | 2009-06-09 | Microsoft Corporation | Matching media file metadata to standardized metadata |
US7818350B2 (en) | 2005-02-28 | 2010-10-19 | Yahoo! Inc. | System and method for creating a collaborative playlist |
JP4577438B2 (en) | 2008-10-21 | 2010-11-10 | ソニー株式会社 | Playlist search device, playlist search method, and program |
US20110295843A1 (en) | 2010-05-26 | 2011-12-01 | Apple Inc. | Dynamic generation of contextually aware playlists |
EP2843860A1 (en) | 2013-08-26 | 2015-03-04 | Panasonic Automotive Systems Company of America, Division of Panasonic Corporation of North America | Method and system for preparing a playlist for an Internet content provider |
US11086936B2 (en) | 2016-12-30 | 2021-08-10 | Spotify Ab | Media content item recommendation system |
US11163817B2 (en) | 2018-05-24 | 2021-11-02 | Spotify Ab | Descriptive media content search |
US20200278997A1 (en) | 2019-03-01 | 2020-09-03 | Spotify Ab | Descriptive media content search from curated content |
-
2018
- 2018-05-24 US US15/988,943 patent/US11163817B2/en active Active
-
2021
- 2021-09-30 US US17/489,937 patent/US11537651B2/en active Active
-
2022
- 2022-11-23 US US17/993,720 patent/US20230205809A1/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040117189A1 (en) * | 1999-11-12 | 2004-06-17 | Bennett Ian M. | Query engine for processing voice based queries including semantic decoding |
US20120303713A1 (en) * | 2001-04-19 | 2012-11-29 | Harbick Andrew V | Playlist-based detection of similar digital works and work creators |
US20050071328A1 (en) * | 2003-09-30 | 2005-03-31 | Lawrence Stephen R. | Personalization of web search |
US20090094231A1 (en) * | 2007-10-05 | 2009-04-09 | Fujitsu Limited | Selecting Tags For A Document By Analyzing Paragraphs Of The Document |
US20160360266A1 (en) * | 2015-06-05 | 2016-12-08 | Google Inc. | Video channel categorization schema |
Also Published As
Publication number | Publication date |
---|---|
US11537651B2 (en) | 2022-12-27 |
US11163817B2 (en) | 2021-11-02 |
US20190361982A1 (en) | 2019-11-28 |
US20220092103A1 (en) | 2022-03-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11537651B2 (en) | Descriptive media content search | |
US9824150B2 (en) | Systems and methods for providing information discovery and retrieval | |
US20200278997A1 (en) | Descriptive media content search from curated content | |
US11636835B2 (en) | Spoken words analyzer | |
US11157542B2 (en) | Systems, methods and computer program products for associating media content having different modalities | |
US20140012859A1 (en) | Personalized dynamic content delivery system | |
US20210335349A1 (en) | Systems and methods for improving fulfillment of media content related requests via utterance-based human-machine interfaces | |
US11762901B2 (en) | User consumption behavior analysis and composer interface | |
US11574627B2 (en) | Masking systems and methods | |
US20190294690A1 (en) | Media content item recommendation system | |
US11653048B2 (en) | Systems and methods for delivering relevant media content by inferring past media content consumption | |
EP3648106B1 (en) | Media content steering | |
US20220188062A1 (en) | Skip behavior analyzer | |
US11886486B2 (en) | Apparatus, systems and methods for providing segues to contextualize media content | |
CN114503099A (en) | Replying to queries with voice recordings |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: SPOTIFY AB, SWEDEN Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:JACOBSON, KURT;LAMERE, PAUL;SIGNING DATES FROM 20181009 TO 20181020;REEL/FRAME:063208/0690 |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: FINAL REJECTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: FINAL REJECTION MAILED |