CN114969514A - Live broadcast recommendation method and device and electronic equipment - Google Patents
Live broadcast recommendation method and device and electronic equipment Download PDFInfo
- Publication number
- CN114969514A CN114969514A CN202210489779.2A CN202210489779A CN114969514A CN 114969514 A CN114969514 A CN 114969514A CN 202210489779 A CN202210489779 A CN 202210489779A CN 114969514 A CN114969514 A CN 114969514A
- Authority
- CN
- China
- Prior art keywords
- user
- live broadcast
- broadcast room
- feature vector
- determining
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 54
- 239000013598 vector Substances 0.000 claims abstract description 169
- 230000004044 response Effects 0.000 claims abstract description 10
- 230000006399 behavior Effects 0.000 claims description 87
- 238000010586 diagram Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 14
- 230000004927 fusion Effects 0.000 claims description 14
- 238000003062 neural network model Methods 0.000 claims description 12
- 238000005516 engineering process Methods 0.000 claims description 10
- 230000007246 mechanism Effects 0.000 claims description 8
- 238000007499 fusion processing Methods 0.000 claims description 3
- 238000013473 artificial intelligence Methods 0.000 abstract description 4
- 238000012545 processing Methods 0.000 description 16
- 238000004891 communication Methods 0.000 description 9
- 230000008569 process Effects 0.000 description 6
- 238000012512 characterization method Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000003993 interaction Effects 0.000 description 3
- 238000005065 mining Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005295 random walk Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 230000000007 visual effect 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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/21—Server components or server architectures
- H04N21/218—Source of audio or video content, e.g. local disk arrays
- H04N21/2187—Live feed
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The disclosure provides a live broadcast recommendation method and device and electronic equipment, relates to the technical field of artificial intelligence, and particularly relates to the technical field of multimedia. The specific implementation scheme is as follows: determining a behavior feature vector of a user based on the behavior of the user; determining a first live broadcast room and a second live broadcast room which meet the similarity condition based on the behavior feature vector of the user; and recommending the second live broadcast room to the first user in response to a live broadcast recommendation request sent by the first user corresponding to the first live broadcast room.
Description
Technical Field
The present disclosure relates to the field of artificial intelligence technologies, and in particular, to a live broadcast recommendation method and apparatus in the field of multimedia technologies, and an electronic device.
Background
With the rapid development of the live broadcast industry, entertainment and product transaction through live broadcast become a life style; in some scenarios, it is desirable to recommend live rooms of interest to the user. Therefore, how to accurately recommend the live broadcast room to the user to improve the click rate of the resources in the live broadcast room is a constantly pursued target in the field of live broadcast recommendation.
Disclosure of Invention
The disclosure provides a live broadcast recommendation method and device and electronic equipment.
According to a first aspect of the present disclosure, a live broadcast recommendation method is provided, including:
determining a behavior feature vector of a user based on the behavior of the user;
determining a first live broadcast room and a second live broadcast room which meet the similarity condition based on the behavior feature vector of the user;
and recommending the second live broadcast room to the first user in response to a live broadcast recommendation request sent by the first user corresponding to the first live broadcast room.
According to a second aspect of the present disclosure, there is provided a live broadcast recommendation apparatus, including:
the behavior feature vector determination module is used for determining a behavior feature vector of a user based on the behavior of the user;
the similar live broadcast room determining module is used for determining a first live broadcast room and a second live broadcast room which meet the similarity condition based on the behavior characteristic vector of the user;
and the recommending module is used for responding to a live broadcast recommending request sent by a first user corresponding to the first live broadcast room and recommending the second live broadcast room to the first user.
According to a third aspect of the present disclosure, there is provided an electronic device comprising:
at least one processor; and a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the live recommendation method described above.
According to a fourth aspect of the present disclosure, there is provided a non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to execute the live recommendation method described above.
According to a fifth aspect of the present disclosure, there is provided a computer program product comprising computer programs/instructions which, when executed by a processor, implement a live recommendation method according to the above.
It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.
Drawings
The drawings are included to provide a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:
fig. 1 is a schematic processing flow diagram of a live broadcast recommendation method in the related art;
fig. 2 is a schematic view of an alternative processing flow of a live broadcast recommendation method provided in an embodiment of the present disclosure;
fig. 3 is a schematic diagram of a specific implementation process for determining a behavior feature vector of a user based on a behavior of the user according to an embodiment of the present disclosure;
fig. 4 is a schematic view of an alternative processing flow for acquiring behavior information of multiple users with respect to a live broadcast room in a first time interval according to an embodiment of the present disclosure;
FIG. 5 is a diagram of behavior of multiple users with respect to a live broadcast room, constructed using graph characterization techniques according to an embodiment of the present disclosure;
fig. 6 is a schematic view of an alternative processing flow for determining that the first live broadcast room and the second live broadcast room satisfy the similarity condition based on the behavior feature vector of the user according to the embodiment of the present disclosure;
fig. 7 is an optional processing flow diagram of recommending a second live broadcast room to a first user in response to a live broadcast recommendation request sent by the first user corresponding to the first live broadcast room, provided by the embodiment of the present disclosure;
FIG. 8 is a schematic diagram of recommending a second live broadcast room to a first user using a neural network model provided by an embodiment of the disclosure;
fig. 9 is a schematic structural diagram of a composition of a live broadcast recommendation apparatus provided in an embodiment of the present disclosure;
fig. 10 is a block diagram of an electronic device used to implement a live recommendation method of an embodiment of the present disclosure.
Detailed Description
Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict.
In the description that follows, references to the terms "first \ second \ third" are intended merely to distinguish similar objects and do not denote a particular order, but rather are to be understood that "first \ second \ third" may, where permissible, be interchanged in a particular order or sequence so that embodiments of the disclosure described herein can be practiced in other than the order shown or described herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The terminology used herein is for the purpose of describing embodiments of the disclosure only and is not intended to be limiting of the disclosure.
Before describing embodiments of the present disclosure in detail, relevant terms related to the present disclosure are explained.
1) And the live broadcast room is used for providing a virtual space for live broadcast of the anchor.
2) Live broadcasting refers to that a user watches videos on different live broadcasting platforms in real time through a network, the content of the videos is generated or uploaded by a main broadcast in real time, and the user can communicate with the main broadcast in real time by sending messages such as barrage and comments, so that the method is a novel social network mode.
As shown in fig. 1, if a user sends a live broadcast recommendation request, a resource aggregation module (Global recommendation, GR) sends a resource request to a live broadcast recommendation queue; the fine ranking module performs fine ranking and scoring on available resources, and acquires user characteristics, resource characteristics and context characteristics from the characteristic module; and reordering the live broadcast room through a plurality of models based on the user characteristics, the resource characteristics and the context characteristics to generate final live broadcast room resources, and feeding back the live broadcast room resources to the resource aggregation module. However, in the live broadcast recommendation method in the related art, features are mined in a manual feature engineering mode, along with continuous mining and construction of the features, difficulty in finding out the features with high relevance and good effect is increased continuously, and labor cost is also greatly increased. In the process of implementing the live broadcast recommendation method, the applicant finds that the importance of the relationship between people in live broadcast recommendation is ignored in the existing live broadcast recommendation method, and the live broadcast recommendation is carried out only by mining user characteristics and resource characteristics, so that the generalization capability of a model is influenced, the recommendation effect on new users and new anchor is poor, and the benign development of live broadcast ecology is influenced.
Fig. 2 is a schematic view of an optional processing flow of a live broadcast recommendation method provided in the present disclosure, where the live broadcast recommendation method may include at least the following steps:
step S201, determining the behavior feature vector of the user based on the behavior of the user.
And S202, determining a first live broadcast room and a second live broadcast room which meet the similarity condition based on the behavior feature vector of the user.
Step S203, recommending the second live broadcast room to the first user in response to a live broadcast recommendation request sent by the first user corresponding to the first live broadcast room.
With respect to step S201, in some optional embodiments, the specific implementation process of determining the behavior feature vector of the user based on the behavior of the user may be as shown in fig. 3, and at least includes:
in step S201a, behavior information of a plurality of users with respect to the live broadcast room in the first time interval is acquired.
Step S201b, determining behavior feature vectors of the multiple users based on the behavior information of the multiple users for the live broadcast room.
As an example, if the duration of the first time interval is one month, behavior information of a plurality of users in a month before the current time for the live broadcast room is acquired.
With respect to step S201a, in some embodiments, the behavior information of the user with respect to the live broadcast room may include: the user clicks on the information of the live room and/or the user effectively watches the information of the live room. The information that the user clicks the live broadcast room may refer to information of the live broadcast room that the user pays attention to in a clicking manner, such as a name of the live broadcast room, an identifier of the live broadcast room, or time that the user clicks the live broadcast room. The information that the user clicks the live broadcast room may also be information of a live broadcast room that the user enters by clicking, such as a name of the live broadcast room, an identifier of the live broadcast room, or time that the user enters the live broadcast room. The effective watching of the live broadcast room by the user can mean that the effective residence time of the user in the live broadcast room is longer than a preset time threshold, and correspondingly, the effective watching information of the user in the live broadcast room can comprise the name of the live broadcast room, the identification of the live broadcast room, the effective watching times of the user in the live broadcast room and the like.
In some embodiments, an optional process flow of acquiring behavior information of a plurality of users with respect to a live broadcast room in a first time interval may be as shown in fig. 4, including:
step a1, constructing a behavior diagram of a plurality of users aiming at the live broadcast room based on a diagram representation technology.
In some embodiments, a behavior graph is built for each user separately for the live room using graph characterization techniques. As an example, if user a clicks on or actively watches live rooms a, b, and d; the user B clicks or effectively watches the live rooms a and c; the user C clicks or effectively watches live broadcasting rooms b and e; user D clicks on or effectively watches live rooms c, D, and e. Then a behavior diagram of multiple users for the live broadcast room, constructed by using the graph characterization technique, may be a bipartite diagram of the live broadcast room of the users, as shown in fig. 5.
In the embodiment of the disclosure, the behavior characteristics of the user are learned by using a graph characterization technology, the user characteristics can be automatically mined from a large number of user behaviors according to the business target, and compared with mining of the user characteristics in a manual characteristic engineering mode in the related technology, the generalization of a live broadcast recommendation model can be increased, and the recommendation effect on new users and new anchor is improved.
Step b1, randomly selecting a path on the behavior diagram, taking a first node on the path as a starting point, querying a second node adjacent to the first node along the path, and so on until reaching a preset path length, thereby obtaining a plurality of groups of nodes.
In some embodiments, the greater the number of paths between two vertices based on FIG. 5, the higher the correlation between the two vertices. The shorter the path length between two vertices, the higher the correlation.
For example, user a is not connected to live rooms c, e, but user a is connected to live room c by 1 path of length 3, and user a is connected to live room e by 2 paths of length 3. Then the correlation between vertices a and e is higher than vertices a and C, so live room e should be ranked before live room C in the recommendation list of user a, because there are two paths between vertices a and e, which are (a, b, C, e) and (a, D, e), respectively. Wherein the out degree of the vertex passed by the (A, b, C, e) path is (3,2,2,2), and the out degree of the vertex passed by the (A, D, D, e) path is (3,2,3, 2). Therefore, (A, D, D, e) passes through a vertex D with a relatively large degree of emittance, so that (A, D, D, e) contributes less to the correlation between vertices A and e than (A, b, C, e).
And c1, constructing graph model samples corresponding to the behavior graphs based on the plurality of groups of nodes.
In some embodiments, steps b and c employ a random walk algorithm survival map model sample.
For step S202b, determining, based on the behavior information of the multiple users for the live broadcast room, a specific implementation procedure of the feature vectors of the multiple users may include: converting unstructured words in the graph model sample into structured feature vectors; determining the structured feature vector as a feature vector of the plurality of users. Wherein, the feature vector of the user can be generated by word2vec technology.
With respect to step S202, in some embodiments, an optional processing flow of determining, based on the behavior feature vector of the user, the first live broadcast room and the second live broadcast room that satisfy the similarity condition is determined, as shown in fig. 6, which at least includes:
and d1, determining two users with cosine similarity meeting the first threshold as similar users based on the behavior feature vectors of the users.
In some embodiments, cosine similarity between behavior feature vectors of any two users is calculated, and two users with cosine similarity greater than or equal to the first threshold are similar users.
In some scenarios, there may be one or more similar users of one user.
And e1, responding to the fact that the number of similar users between the first live broadcast room and the second live broadcast room is larger than a second threshold value, and determining that the first live broadcast room and the second live broadcast room meet a similarity condition.
In some embodiments, the number of similar users in the users corresponding to any two live broadcast rooms is calculated, and if the number of similar users in the two live broadcast rooms is greater than a second threshold value, it is determined that the two live broadcast rooms meet the similarity condition.
As an example, if the number of users in the first live broadcast is M, the number of users in the second live broadcast is N; and if the number of the similar users in the first live broadcast room and the second live broadcast room is X and X is greater than a preset second threshold value, determining that the first live broadcast room and the second live broadcast room meet the similarity condition.
As for step S203, in response to a live broadcast recommendation request sent by a first user corresponding to a first live broadcast room, an optional processing flow for recommending a second live broadcast room to the first user may at least include, as shown in fig. 7:
step f1, determining a first fused feature vector of the first user's portrait feature vector and the plurality of anchor portrait feature vectors corresponding to the first user.
In specific implementation, the portrait feature vector of the first user and the portrait feature vectors of the multiple anchor corresponding to the first user may be determined; inputting the image features of the plurality of anchor into a self-attention mechanism neural network model, and determining a first direct-playing image feature vector based on the output of the self-attention mechanism neural network model; inputting the portrait feature vector of the first user and the portrait feature vectors of the plurality of anchor into a deep interest network model, and determining candidate fusion feature vectors based on an output of the deep interest network model; and performing fusion processing on the first live image feature vector, the image feature vector of the first user and the candidate fusion feature vector to obtain a first fusion feature vector of the image feature of the first user and the image features of a plurality of anchor corresponding to the first user.
The image feature vector of the first user can be used for representing the gender, age, occupation and the like of the first user; the portrait feature vectors of the anchor may be used to characterize the gender, age, occupation, content and characteristics of the anchor within the live room, and the like. The portrait feature vector of the first user and the portrait feature vector of the anchor may be gcf vectors.
In some embodiments, the self-attention mechanism neural network model and the deep interest network model are both a priori neural network models.
Step g1, determining a second fusion feature vector of the behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user.
The resource characteristic vectors of the live broadcast rooms associated with the first user can be used for representing the opening time of the live broadcast room, the online number of people in the live broadcast room, the activity of the live broadcast room and the like; the liveness of the live broadcast room can be interaction between a main broadcast and a user, praise of the user for the main broadcast and the like; resource feature vectors of a plurality of live rooms associated with the first user may be obtained through a real-time interface.
In some embodiments, in determining the second fused feature vector, a cross vector between the behavior feature vector of the first user and the resource feature vectors of the plurality of live rooms associated with the first user may be determined based on the behavior feature vector of the first user and the resource feature vectors of the plurality of live rooms associated with the first user; determining the second fused feature vector based on the behavior feature vector of the first user, the resource feature vectors of the plurality of live rooms associated with the first user, and the cross vector. Wherein a cross vector between the behavior feature vector of the first user and the resource feature vectors of the plurality of live rooms associated with the first user may be: the number of times the first user clicks the live broadcast room historically, the duration of time the first user watches the live broadcast room historically, and the like. The cross vector is used for representing cross characteristics between the first user and the live broadcast rooms, can be determined based on the behavior characteristic vector of the first user and resource characteristic vectors of a plurality of live broadcast rooms associated with the first user, and can also be obtained through a real-time interface. The second fused feature vector is a feature vector obtained after multi-layer neural network (MLP) dimensionality reduction processing, so that the second fused feature vector has the same dimensionality as the first fused feature vector.
Step h1, determining the second live broadcast room based on the first fused feature vector and the second fused feature vector.
In some embodiments, the first fused feature vector and the second fused feature vector may be input into a live broadcast room determination model after being subjected to a splicing process; and determining the output of the live broadcast room determination model as a second live broadcast room.
The live broadcast room determination model is a priori neural network model, and can determine a live broadcast room recommended to a user according to input user information and live broadcast information (such as anchor information or live broadcast room information).
Step i1, recommending the second live broadcast room to the first user.
Based on the description of fig. 7, a schematic diagram of recommending a second live broadcast room to a first user by using a neural network model may be as shown in fig. 8, where an image feature vector of the first user is represented by usergcf, an image feature vector of a anchor is represented by seedgcf1, seedgcf2.
In the embodiment of the disclosure, the vector characteristics of the user are processed through the self-attention mechanism neural network model and the deep interest network model, so that effective information can be effectively extracted from the user vector, and the efficiency and the matching degree of recommending a live broadcast room for the user are improved.
According to the live broadcast recommendation method provided by the embodiment of the disclosure, the first live broadcast room and the second live broadcast room which meet the similarity condition are determined according to the behavior feature vector of the user, so that the association relationship between the live broadcast rooms is established by taking the relationship between the users as a dimension. Because the first live broadcast room and the second live broadcast room are established based on the relationship between users, the users between the first live broadcast room and the second live broadcast room have similarity, and the first live broadcast room and the second live broadcast room also have similarity; when a first user of a first live broadcast room sends a live broadcast recommendation request, a second live broadcast room is recommended to the first user, the accuracy of live broadcast recommendation can be improved, the live broadcast room is accurately recommended for the user, and the click rate of resources in the live broadcast room is improved.
The embodiment of the present disclosure further provides a live broadcast recommendation apparatus, a structure of the live broadcast recommendation apparatus is shown in fig. 9, and includes:
a behavior feature vector determination module 301, configured to determine a behavior feature vector of a user based on a behavior of the user;
a similar live broadcast room determining module 302, configured to determine, based on the behavior feature vector of the user, a first live broadcast room and a second live broadcast room that meet a similarity condition;
and a live broadcast room recommending module 303, configured to recommend the second live broadcast room to the first user in response to a live broadcast recommendation request sent by the first user corresponding to the first live broadcast room.
In some optional embodiments, the behavior feature vector determining module 301 is configured to obtain behavior information of a plurality of users for a live broadcast room in a first time interval;
and determining the behavior characteristic vectors of the plurality of users based on the behavior information of the plurality of users for the live broadcast room.
In some optional embodiments, the behavior feature vector determination module 301 is configured to construct a behavior diagram of a plurality of users for a live broadcast room by using a graph characterization technique;
randomly selecting a path on the behavior diagram, taking a first node on the path as a starting point, inquiring a second node adjacent to the first node along the path, and so on until a preset path length is reached to obtain a plurality of groups of nodes;
and constructing graph model samples corresponding to the behavior graphs on the basis of the plurality of groups of nodes.
In some optional embodiments, the behavior feature vector determination module 301 is configured to convert an unstructured word in the graph model sample into a structured feature vector;
determining the structured feature vector as a feature vector of the plurality of users.
In some optional embodiments, the behavior information of the plurality of users with respect to the live broadcast room in the first time interval includes:
each user in the plurality of users clicks on information of a live broadcast room in the first time interval; and/or each user in the plurality of users effectively watches the information of the live broadcast room in the first time.
In some optional embodiments, the live room recommendation module 303 is configured to determine a first blended feature vector of the portrait feature vector of the first user and portrait feature vectors of a plurality of anchor corresponding to the first user;
determining a second fused feature vector of the behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user;
determining the second live broadcast room based on the first fused feature vector and the second fused feature vector; recommending the second live broadcast room to the first user.
In some optional embodiments, the live broadcast recommendation module 303 is configured to determine a portrait feature vector of the first user and portrait feature vectors of multiple anchor casts corresponding to the first user;
inputting the plurality of anchor portrait features into a self-attention mechanism neural network model, and determining a first direct-playing portrait feature vector based on an output of the self-attention mechanism neural network model;
inputting the portrait feature vector of the first user and the portrait feature vectors of the plurality of anchor into a deep interest network model, and determining candidate fusion feature vectors based on the output of the deep interest network model;
and performing fusion processing on the first live image feature vector, the image feature vector of the first user and the candidate fusion feature vector to obtain a first fusion feature vector of the image feature of the first user and the image features of a plurality of anchor corresponding to the first user.
In some optional embodiments, the live broadcast room recommending module 303 is configured to determine a behavior feature vector of the first user and resource feature vectors of multiple live broadcast rooms associated with the first user;
determining a cross vector between the behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user based on the behavior feature vector of the first user and the resource feature vectors of the plurality of live rooms associated with the first user;
determining the second fused feature vector based on the behavior feature vector of the first user, the resource feature vectors of the plurality of live rooms associated with the first user, and the cross vector.
In some optional embodiments, the live broadcast room recommending module 303 is configured to input a live broadcast room determination model after the first fused feature vector and the second fused feature vector are spliced;
determining an output of the live room determination model as the second live room.
In some optional embodiments, the similar live broadcast room determining module 302 is configured to determine, based on the behavior feature vector of the user, that two users whose cosine similarity satisfies a first threshold are similar users;
and determining that the first live broadcast room and the second live broadcast room meet a similarity condition in response to the number of similar users between the first live broadcast room and the second live broadcast room being greater than a second threshold.
It should be noted that, in the technical solution of the present disclosure, the acquisition, storage, application, and the like of the personal information of the related user all meet the regulations of the relevant laws and regulations, and do not violate the common customs of the public order.
The present disclosure also provides an electronic device, a readable storage medium, and a computer program product according to embodiments of the present disclosure.
FIG. 10 shows a schematic block diagram of an example electronic device 800 that may be used to implement embodiments of the present disclosure. In some alternative embodiments, the electronic device 800 may be a terminal device or a server. In some alternative embodiments, the electronic device 800 may implement the live broadcast recommendation method provided by the embodiment of the present application by running a computer program, for example, the computer program may be a native program or a software module in an operating system; may be a local (Native) Application (APP), i.e. a program that needs to be installed in the operating system to run; or may be an applet, i.e. a program that can be run only by downloading it to the browser environment; but also an applet that can be embedded into any APP. In general, the computer programs described above may be any form of application, module or plug-in.
In practical applications, the electronic device 800 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a Cloud server providing basic Cloud computing services such as a Cloud service, a Cloud database, Cloud computing, a Cloud function, Cloud storage, a network service, Cloud communication, a middleware service, a domain name service, a security service, a CDN, and a big data and artificial intelligence platform, where Cloud Technology (Cloud Technology) refers to a hosting Technology for unifying series resources such as hardware, software, and a network in a wide area network or a local area network to implement computing, storage, processing, and sharing of data. The electronic device 800 may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart television, a smart watch, and the like.
Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, in-vehicle terminals, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the disclosure described and/or claimed herein.
As shown in fig. 10, the electronic device 800 includes a computing unit 801 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM)802 or a computer program loaded from a storage unit 808 into a Random Access Memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the electronic apparatus 800 can also be stored. The calculation unit 801, the ROM802, and the RAM 803 are connected to each other by a bus 804. An input/output (I/O) interface 805 is also connected to bus 804.
A number of components in the electronic device 800 are connected to the I/O interface 805, including: an input unit 806 such as a keyboard, a mouse, or the like; an output unit 807 such as various types of displays, speakers, and the like; a storage unit 808, such as a magnetic disk, optical disk, or the like; and a communication unit 809 such as a network card, modem, wireless communication transceiver, etc. The communication unit 809 allows the electronic device 800 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunication networks.
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.
Program code for implementing the live recommendation methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.
The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server with a combined blockchain.
It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present disclosure may be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and the present disclosure is not limited herein.
The above detailed description should not be construed as limiting the scope of the disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present disclosure should be included in the scope of protection of the present disclosure.
Claims (14)
1. A live recommendation method includes:
determining a behavior feature vector of a user based on the behavior of the user;
determining a first live broadcast room and a second live broadcast room which meet the similarity condition based on the behavior feature vector of the user;
and recommending the second live broadcast room to the first user in response to a live broadcast recommendation request sent by the first user corresponding to the first live broadcast room.
2. The method of claim 1, wherein the determining a behavior feature vector of the user based on the user's behavior comprises:
acquiring behavior information of a plurality of users aiming at a live broadcast room in a first time interval;
and determining the behavior characteristic vectors of the plurality of users based on the behavior information of the plurality of users for the live broadcast room.
3. The method of claim 2, wherein the obtaining behavior information of a plurality of users with respect to a live broadcast room in a first time interval comprises:
constructing a behavior diagram of a plurality of users for a live broadcast room by using a graph representation technology;
randomly selecting a path on the behavior diagram, taking a first node on the path as a starting point, inquiring a second node adjacent to the first node along the path, and so on until a preset path length is reached to obtain a plurality of groups of nodes;
and constructing graph model samples corresponding to the behavior graphs on the basis of the plurality of groups of nodes.
4. The method of claim 3, wherein the determining feature vectors of the plurality of users based on behavior information of the plurality of users for a live room comprises:
converting unstructured words in the graph model sample into structured feature vectors;
determining the structured feature vector as a feature vector of the plurality of users.
5. The method of any of claims 2 to 4, wherein the behavior information of the plurality of users with respect to the live broadcast room in the first time interval comprises:
each user in the plurality of users clicks on information of a live broadcast room in the first time interval;
and/or each user in the plurality of users effectively watches the information of the live broadcast room in the first time.
6. The method of claim 1, wherein the recommending the second live broadcast room to the first user in response to a live broadcast recommendation request sent by a first user corresponding to the first live broadcast room comprises:
determining a first fused feature vector of the portrait feature vector of the first user and portrait feature vectors of a plurality of anchor corresponding to the first user;
determining a second fused feature vector of the behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user;
determining the second live broadcast room based on the first fused feature vector and the second fused feature vector;
recommending the second live broadcast room to the first user.
7. The method of claim 6, wherein said determining a first blended feature vector of the portrait feature vector of the first user and a plurality of anchor portrait feature vectors corresponding to the first user comprises:
determining the portrait feature vector of the first user and portrait feature vectors of a plurality of anchor characters corresponding to the first user;
inputting the plurality of anchor portrait features into a self-attention mechanism neural network model, and determining a first anchor portrait feature vector based on an output of the self-attention mechanism neural network model;
inputting the portrait feature vector of the first user and the portrait feature vectors of the plurality of anchor into a deep interest network model, and determining candidate fusion feature vectors based on the output of the deep interest network model;
and performing fusion processing on the first live image feature vector, the image feature vector of the first user and the candidate fusion feature vector to obtain a first fusion feature vector of the image feature of the first user and the image features of a plurality of anchor corresponding to the first user.
8. The method of claim 6, wherein the determining a second fused feature vector of the behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user comprises:
determining a behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user;
determining a cross vector between the behavior feature vector of the first user and resource feature vectors of a plurality of live rooms associated with the first user based on the behavior feature vector of the first user and the resource feature vectors of the plurality of live rooms associated with the first user;
determining the second fused feature vector based on the behavior feature vector of the first user, the resource feature vectors of the plurality of live rooms associated with the first user, and the cross vector.
9. The method of claim 6, wherein the determining the second live broadcast room based on the first fused feature vector and the second fused feature vector comprises:
after splicing the first fusion characteristic vector and the second fusion characteristic vector, inputting the first fusion characteristic vector and the second fusion characteristic vector into a live broadcast room to determine a model;
determining an output of the live room determination model as the second live room.
10. The method of claim 1, wherein determining, based on the behavior feature vector of the user, a first live broadcast room and a second live broadcast room that satisfy a similarity condition comprises:
determining two users with cosine similarity meeting a first threshold as similar users based on the behavior feature vectors of the users;
and determining that the first live broadcast room and the second live broadcast room meet the similarity condition in response to the number of similar users between the first live broadcast room and the second live broadcast room being larger than a second threshold value.
11. A live recommendation device, the live recommendation device comprising:
the behavior feature vector determination module is used for determining a behavior feature vector of a user based on the behavior of the user;
the similar live broadcast room determining module is used for determining a first live broadcast room and a second live broadcast room which meet the similarity condition based on the behavior characteristic vector of the user;
and the live broadcast room recommending module is used for responding to a live broadcast recommending request sent by a first user corresponding to the first live broadcast room and recommending the second live broadcast room to the first user.
12. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 10.
13. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1 to 10.
14. A computer program product comprising a computer program/instructions which, when executed by a processor, implement the method of any one of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210489779.2A CN114969514B (en) | 2022-05-06 | 2022-05-06 | Live broadcast recommendation method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210489779.2A CN114969514B (en) | 2022-05-06 | 2022-05-06 | Live broadcast recommendation method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114969514A true CN114969514A (en) | 2022-08-30 |
| CN114969514B CN114969514B (en) | 2024-10-15 |
Family
ID=82981881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210489779.2A Active CN114969514B (en) | 2022-05-06 | 2022-05-06 | Live broadcast recommendation method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114969514B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106560811A (en) * | 2016-09-23 | 2017-04-12 | 武汉斗鱼网络科技有限公司 | Direct broadcasting room recommending method and system based on broadcaster style |
| CN107613395A (en) * | 2017-08-28 | 2018-01-19 | 武汉斗鱼网络科技有限公司 | Recommend method and system in live room |
| CN107835441A (en) * | 2017-10-10 | 2018-03-23 | 武汉斗鱼网络科技有限公司 | Live recommendation method, storage medium, equipment and system based on path prediction |
| CN108536814A (en) * | 2018-04-04 | 2018-09-14 | 武汉斗鱼网络科技有限公司 | Direct broadcasting room recommends method, computer readable storage medium and electronic equipment |
| CN108712663A (en) * | 2018-05-03 | 2018-10-26 | 武汉斗鱼网络科技有限公司 | Direct broadcasting room based on bipartite graph recommends method, corresponding medium and equipment |
| CN109862013A (en) * | 2019-01-31 | 2019-06-07 | 武汉斗鱼网络科技有限公司 | A kind of direct broadcasting room recommended method, storage medium, electronic equipment and system |
| US20190325401A1 (en) * | 2017-01-31 | 2019-10-24 | Sony Corporation | Information processing apparatus, information processing method, and program |
| CN113301442A (en) * | 2021-05-20 | 2021-08-24 | 北京百度网讯科技有限公司 | Method, apparatus, medium, and program product for determining live broadcast resource |
| CN114040216A (en) * | 2021-11-03 | 2022-02-11 | 杭州网易云音乐科技有限公司 | Live broadcast room recommendation method, medium, device and computing equipment |
| CN114302152A (en) * | 2021-11-17 | 2022-04-08 | 北京乐我无限科技有限责任公司 | Live broadcast room recommendation method, device, equipment and storage medium |
-
2022
- 2022-05-06 CN CN202210489779.2A patent/CN114969514B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106560811A (en) * | 2016-09-23 | 2017-04-12 | 武汉斗鱼网络科技有限公司 | Direct broadcasting room recommending method and system based on broadcaster style |
| US20190325401A1 (en) * | 2017-01-31 | 2019-10-24 | Sony Corporation | Information processing apparatus, information processing method, and program |
| CN107613395A (en) * | 2017-08-28 | 2018-01-19 | 武汉斗鱼网络科技有限公司 | Recommend method and system in live room |
| CN107835441A (en) * | 2017-10-10 | 2018-03-23 | 武汉斗鱼网络科技有限公司 | Live recommendation method, storage medium, equipment and system based on path prediction |
| CN108536814A (en) * | 2018-04-04 | 2018-09-14 | 武汉斗鱼网络科技有限公司 | Direct broadcasting room recommends method, computer readable storage medium and electronic equipment |
| CN108712663A (en) * | 2018-05-03 | 2018-10-26 | 武汉斗鱼网络科技有限公司 | Direct broadcasting room based on bipartite graph recommends method, corresponding medium and equipment |
| CN109862013A (en) * | 2019-01-31 | 2019-06-07 | 武汉斗鱼网络科技有限公司 | A kind of direct broadcasting room recommended method, storage medium, electronic equipment and system |
| CN113301442A (en) * | 2021-05-20 | 2021-08-24 | 北京百度网讯科技有限公司 | Method, apparatus, medium, and program product for determining live broadcast resource |
| CN114040216A (en) * | 2021-11-03 | 2022-02-11 | 杭州网易云音乐科技有限公司 | Live broadcast room recommendation method, medium, device and computing equipment |
| CN114302152A (en) * | 2021-11-17 | 2022-04-08 | 北京乐我无限科技有限责任公司 | Live broadcast room recommendation method, device, equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 王硕慜;尹菲;刘增涛;: "基于广电大数据分析的用户画像与个性化推荐研究", 有线电视技术, no. 08, 15 August 2018 (2018-08-15) * |
| 郭淑慧;吕欣;: "网络直播平台数据挖掘与行为分析综述", 物理学报, no. 08, 23 April 2020 (2020-04-23) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114969514B (en) | 2024-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108520470B (en) | Method and apparatus for generating user attribute information | |
| CN114265979B (en) | Method for determining fusion parameters, information recommendation method and model training method | |
| CN109471978B (en) | Electronic resource recommendation method and device | |
| JP7498248B2 (en) | Content recommendation and sorting model training method, device, equipment, storage medium, and computer program | |
| CN114357197B (en) | Event reasoning method and device | |
| CN112989146A (en) | Method, apparatus, device, medium, and program product for recommending resources to a target user | |
| CN114119123A (en) | Information pushing method and device | |
| CN113792212A (en) | Multimedia resource recommendation method, device, equipment and storage medium | |
| CN113365090B (en) | Object recommendation method, object recommendation device, electronic equipment and readable storage medium | |
| CN113961797A (en) | Resource recommendation method and device, electronic equipment and readable storage medium | |
| CN116597443A (en) | Material tag processing method and device, electronic equipment and medium | |
| CN113590447B (en) | Buried point processing method and device | |
| CN115168732A (en) | Resource recommendation method, device, equipment and storage medium | |
| CN115619142A (en) | Method, device, equipment and computer readable medium for matching data | |
| CN114969514B (en) | Live broadcast recommendation method and device and electronic equipment | |
| CN114037060A (en) | Pre-training model generation method and device, electronic equipment and storage medium | |
| CN112632384A (en) | Data processing method and device for application program, electronic equipment and medium | |
| CN110942306A (en) | Data processing method and device and electronic equipment | |
| CN113691442B (en) | Friend recommendation method, device, equipment, storage medium and program product | |
| CN113572679B (en) | Account intimacy generation method and device, electronic equipment and storage medium | |
| CN113360756A (en) | Resource recommendation method and device, electronic equipment and medium | |
| CN117851653A (en) | Object matching method, device, electronic equipment, storage medium and program product | |
| CN118035557A (en) | Resource pushing method, device, electronic equipment, storage medium and program product | |
| CN114185859A (en) | File processing method and device and electronic equipment | |
| CN113343090A (en) | Method, apparatus, device, medium and product for pushing information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |