CN110598618A - Content recommendation method and device, computer equipment and computer-readable storage medium - Google Patents

Content recommendation method and device, computer equipment and computer-readable storage medium Download PDF

Info

Publication number
CN110598618A
CN110598618A CN201910838291.4A CN201910838291A CN110598618A CN 110598618 A CN110598618 A CN 110598618A CN 201910838291 A CN201910838291 A CN 201910838291A CN 110598618 A CN110598618 A CN 110598618A
Authority
CN
China
Prior art keywords
videos
recommendation
video
user
characteristic information
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
Application number
CN201910838291.4A
Other languages
Chinese (zh)
Inventor
杨伟俊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tencent Technology Shenzhen Co Ltd
Original Assignee
Tencent Technology Shenzhen Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tencent Technology Shenzhen Co Ltd filed Critical Tencent Technology Shenzhen Co Ltd
Priority to CN201910838291.4A priority Critical patent/CN110598618A/en
Publication of CN110598618A publication Critical patent/CN110598618A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Multimedia (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The embodiment of the application discloses a content recommendation method, a content recommendation device, computer equipment and a computer readable storage medium; acquiring a plurality of videos historically watched by a user; extracting image frames from the videos to obtain an image frame set of each video; extracting multi-dimensional characteristic information of a target object from an image frame set of each video to obtain the multi-dimensional characteristic information of each video; calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos; fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set; and sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set. The scheme can improve the accuracy of content recommendation.

Description

Content recommendation method and device, computer equipment and computer-readable storage medium
Technical Field
The present application relates to the field of communications technologies, and in particular, to a content recommendation method and apparatus, a computer device, and a computer-readable storage medium.
Background
With the development of communication technology, online shopping is becoming an integral part of life, and the number and types of commodities are rapidly increasing nowadays, so that a user needs to spend a lot of time to find a desired commodity when shopping online.
In the research and practice process of the related technology, the inventor of the present application finds that the content that the user is interested in, such as commodity information, can be recommended to the user according to the interest characteristics and historical purchasing behavior of the user, and the content recommended to the user by the existing video website generally needs to acquire the operation record data of the user on various shopping websites, such as the related data of the attention situation and the historical purchasing record, or social data, but the content recommended to the user is not accurate.
Disclosure of Invention
The embodiment of the application provides a content recommendation method, a content recommendation device, computer equipment and a computer readable storage medium, which can improve the accuracy of content recommendation.
The embodiment of the application provides a content recommendation method, which comprises the following steps:
acquiring a plurality of videos historically watched by a user;
extracting image frames from the videos to obtain an image frame set of each video;
extracting multi-dimensional characteristic information of a target object from an image frame set of each video to obtain the multi-dimensional characteristic information of each video;
calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos;
fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set;
and sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
Correspondingly, an embodiment of the present application provides a content recommendation device, including:
a first acquisition unit configured to acquire a plurality of videos historically viewed by a user;
the first extraction unit is used for extracting image frames from the videos to obtain an image frame set of each video;
the second extraction unit is used for extracting multi-dimensional characteristic information of the target object from the image frame set of each video to obtain the multi-dimensional characteristic information of each video;
the first calculation unit is used for calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos;
the fusion unit is used for fusing the recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set;
and the sending unit is used for sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
In one embodiment, the first computing unit includes:
the first acquisition subunit is used for acquiring the number of target image frames corresponding to each dimension characteristic information from the image frame set of each video;
and the first calculating subunit is configured to calculate, based on the number of target image frames and the number of image frames in the image frame set, a recommended parameter corresponding to each piece of dimensional feature information, so as to obtain a recommended parameter set corresponding to the plurality of videos.
In one embodiment, the first computing unit includes:
the second obtaining subunit is used for obtaining a preset weight parameter set corresponding to each dimension characteristic information;
and the second calculating subunit is configured to calculate, according to the preset weight parameter set and the video attribute information corresponding to the multiple videos, a recommended parameter corresponding to each dimension feature information to obtain recommended parameter sets corresponding to the multiple videos.
In an embodiment, the first computing subunit further includes:
and the third acquisition subunit is used for acquiring the video types of the plurality of videos and counting the number of each video type.
In one embodiment, the first computing subunit includes:
and the third calculation subunit is used for calculating the recommended parameters corresponding to each dimension characteristic information according to the preset weight parameter set and the number of each video type to obtain recommended parameter sets corresponding to the plurality of videos.
In an embodiment, the first computing subunit further includes:
and the selecting subunit is used for selecting videos matched with a preset video from the videos to obtain a matched video set, and counting the number of the videos in the matched video set.
In one embodiment, the first computing subunit includes:
and the fourth calculating subunit is configured to calculate, based on the preset weight parameter set and the number of videos in the matching video set, a recommendation parameter corresponding to each piece of dimensional feature information to obtain recommendation parameter sets corresponding to the multiple videos.
In an embodiment, the content recommendation apparatus may further include:
a second obtaining unit, configured to obtain a second user set similar to the video watched by the user history based on the videos watched by the user history;
a third obtaining unit, configured to obtain a second viewing video set of the second user set;
and the second calculating unit is used for calculating similar recommendation parameters corresponding to each dimension characteristic information based on the plurality of videos and the second watching video set to obtain similar recommendation parameter sets corresponding to the plurality of videos.
In one embodiment, the second computing unit includes:
a counting subunit, configured to count, based on the plurality of videos and the second viewing video set, the number of historical viewing videos corresponding to the user and each second user;
a fifth calculating subunit, configured to calculate, according to the number of the historical viewing videos, similarities between the user and the videos watched by the second users, so as to obtain a set of similarity between the videos watched by the user;
and the sixth calculating subunit is configured to calculate, according to the video-watching similarity set, a similar recommendation parameter corresponding to each dimension feature information, so as to obtain a similar recommendation parameter set corresponding to the multiple videos.
In one embodiment, the sending unit includes:
the fusion subunit is used for fusing the target recommendation parameter set and the similar recommendation parameter set to obtain a fused target recommendation parameter set;
and the sending subunit is used for sending the recommended content to the terminal of the user according to the fused target recommended parameter set.
Accordingly, embodiments of the present application further provide a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes steps in the content recommendation method provided in any of the embodiments of the present application.
Correspondingly, an embodiment of the present application further provides a computer-readable storage medium, where the storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to perform steps in the content recommendation method provided in any of the embodiments of the present application.
The method comprises the steps of acquiring a plurality of videos historically watched by a user; then, extracting image frames from the videos to obtain an image frame set of each video; extracting multi-dimensional characteristic information of the target object from the image frame set of each video to obtain the multi-dimensional characteristic information of each video; calculating a recommendation parameter corresponding to each dimension characteristic information of each video, and then obtaining a recommendation parameter set of each video, or certainly obtaining recommendation parameter sets corresponding to the plurality of videos; fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set; and finally, sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set. According to the scheme, the original data of the content recommendation, which needs to be acquired, can be data in a video website, and the accuracy of the content recommendation can be improved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a scene schematic diagram of a content recommendation method provided in an embodiment of the present application;
FIG. 2 is a flowchart of a content recommendation method provided in an embodiment of the present application;
FIG. 3 is another flowchart of a content recommendation method provided by an embodiment of the present application;
fig. 4 is a schematic structural diagram of a content recommendation device provided in an embodiment of the present application;
fig. 5 is a schematic structural diagram of a computer device provided in an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The embodiment of the application provides a content recommendation method, a content recommendation device, computer equipment and a computer readable storage medium. Specifically, the embodiment of the present application provides a content recommendation apparatus suitable for a computer device, where the computer device may be a network side device such as a server.
In the embodiment of the present application, a computer device is taken as an example to introduce a content recommendation method. Referring to fig. 1, the server may acquire a plurality of videos that a user has historically viewed; then, extracting image frames from the video to obtain an image frame set of each video; extracting multi-dimensional characteristic information of the target object from the image frame set of each video, wherein the characteristic information of the target object has multiple dimensions, and obtaining the multi-dimensional characteristic information of each video; calculating a recommendation parameter corresponding to each dimension characteristic information of each video, wherein the recommendation parameter corresponding to each dimension characteristic information of each video can be calculated at the moment, and then a recommendation parameter set corresponding to the plurality of videos can be obtained through the recommendation parameters; fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set; and finally, sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
As can be seen from the above, the original data of the content recommendation that needs to be acquired in this embodiment may be data in a video website, and the accuracy of the content recommendation may be improved.
The following are detailed below, and it should be noted that the order of description of the following examples is not intended to limit the preferred order of the examples.
The embodiments of the present application will be described from the perspective of a content recommendation device, which may be specifically integrated in a server.
An embodiment of the present application provides a content recommendation method, which may be executed by a processor of a server, and as shown in fig. 2, a specific flow of the content recommendation method may be as follows:
101. the method comprises the steps of obtaining a plurality of videos historically watched by a user, and extracting image frames from the videos to obtain an image frame set of each video.
The videos watched by the user in the history, that is, the videos watched before currently, may be obtained as the videos watched by the user in the history, where the videos watched in a certain history time before currently may include different types of videos, such as videos of comedy type, videos of literature type, videos of popular types, such as videos of popular names recently, and the like.
The image frames are extracted from a plurality of videos, for example, the image frames may be set at fixed time intervals to extract the videos, or the image frames of the videos may be set at different time intervals to extract the image frames of the videos, and the image frame set of each video may be obtained by aggregating the image frames extracted from each video, and of course, the image frame sets of the videos may also be obtained by the image frame set of each video.
102. And extracting multi-dimensional characteristic information of the target object from the image frame set of each video to obtain the multi-dimensional characteristic information of each video.
For example, the target object is a person, and the characteristic information of the person in different dimensions or angles may be color characteristic information of clothing worn, style characteristic information of clothing, and the like, where the person is not limited to the same person, but refers to all persons in the image frame set.
The image frame set of each video includes multi-dimensional feature information of the target object, where the multi-dimensional feature information may be feature information in the image frames of the video, for example, the extracted image frame set of the video includes color feature information, style feature information of people's clothing, and the like, and the color feature information is feature information of one dimension, and the style feature information is feature information of one dimension.
For example, color feature information of a target object is extracted from an image frame set acquired from a plurality of videos, the color feature information includes first color feature information, second color feature information, and third color feature information, and so on, a color value set corresponding to the color feature information is acquired through image recognition, and if the color values do not match the color values in the color database, the acquired special color values are converted into color values in the color database.
103. And calculating the recommendation parameters corresponding to the dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos.
In one embodiment, the recommendation parameter may be referred to as a preference parameter, and the higher the value of the recommendation parameter is, the deeper the preference degree of the dimension is, for example, taking the target object as an example, each dimension characteristic information of each video target object is calculated, so that the style characteristic information of the clothes worn by the person may be any style, and the recommendation parameter corresponding to the style characteristic information is calculated by the number of extracted image frames to indicate the preference degree of the user to a certain style of the clothes, and the styles may include shirts, dresses, jeans, and the like. For another example, in the feature information of the color dimension in step 102, the recommended parameter value corresponding to the first color feature information is the highest, and the other dimensions have a similar recommended parameter, respectively, so that the recommended parameters are combined into the recommended parameter set of each video.
Optionally, in an embodiment, the step of "calculating a recommended parameter corresponding to each dimension feature information of each video, and obtaining a recommended parameter set corresponding to the plurality of videos" includes:
acquiring the number of target image frames corresponding to each dimension characteristic information from the image frame set of each video;
and calculating a recommendation parameter corresponding to each dimension characteristic information based on the number of the target image frames and the number of the image frames of the image frame set to obtain a recommendation parameter set corresponding to the plurality of videos.
For example, the plurality of videos historically viewed by the user include a first video and a second video, image frames are extracted from the first video and the second video, image frame sets of the first video and the second video are obtained, and feature information of multiple dimensions of the target object, such as color feature information, style feature information, and the like, is extracted from the image frame sets of the two videos.
For the color feature information, including the first color feature information, the second color feature information, and the third color feature information, and so on, the number of color image frames corresponding to the color feature information, such as the number of first color image frames corresponding to the first color feature information, the number of second color image frames corresponding to the second color feature information, and the number of third color image frames corresponding to the third color feature information, is obtained from the image frame sets of the two videos, and the recommended parameter corresponding to the color feature information may be calculated by obtaining the number of color image frames corresponding to each color feature information and the total number of color image frames corresponding to all color feature information.
Similarly, recommendation parameters corresponding to feature information of other dimensions can also be calculated, and the recommendation parameters form a recommendation parameter set of each video.
Optionally, in an embodiment, the step of "calculating a recommended parameter corresponding to each dimension feature information of each video, and obtaining a recommended parameter set corresponding to the plurality of videos" includes:
acquiring a preset weight parameter set corresponding to each dimension characteristic information;
and calculating a recommendation parameter corresponding to each dimension characteristic information according to the preset weight parameter set and the video attribute information corresponding to the videos to obtain a recommendation parameter set corresponding to the videos.
Alternatively, the video attribute information may include type information of the video, specific video information, and the like.
Optionally, in an embodiment, the method for obtaining the recommendation parameter sets corresponding to the multiple videos further includes:
acquiring the video types of the plurality of videos, and counting the number of each video type;
and calculating the recommendation parameter corresponding to each dimension characteristic information according to the preset weight parameter set and the number of each video type to obtain recommendation parameter sets corresponding to the plurality of videos.
For example, the plurality of videos historically viewed by the user include a plurality of types of videos, such as a comedy type video, an art type video, and the like, an image frame is extracted from the comedy type video and the art type video, an image frame set of the comedy type video and the art type video is obtained, and feature information of a plurality of dimensions of the target object, such as color feature information, style feature information, and the like, is extracted from the image frame sets of the two videos.
The style characteristic information comprises first style characteristic information, second style characteristic information, third style characteristic information and the like, a preset type weight parameter set corresponding to the style characteristic information is acquired from image frame sets of the two videos, and recommendation parameters corresponding to the style characteristic information can be calculated through the acquired preset type weight parameter corresponding to each style characteristic information, the number of watched comedy-type videos, the number of literature-type videos and the total watching number of the comedy-type videos and the literature-type videos.
Similarly, recommendation parameters corresponding to feature information of other dimensions can also be calculated, and the recommendation parameters form a recommendation parameter set of each video.
Optionally, in an embodiment, the method for obtaining the recommendation parameter sets corresponding to the multiple videos further includes:
selecting videos matched with a preset video from the videos to obtain a matched video set, and counting the number of the videos of the matched video set;
and calculating recommendation parameters corresponding to each dimension characteristic information based on the preset weight parameter set and the number of videos in the matched video set to obtain recommendation parameter sets corresponding to the videos.
The matched video may be understood as a video of the same video as the preset video in the plurality of videos, and may be a single video or a video set.
For example, a preset video set may be obtained, videos in the preset videos watched by the user are obtained based on the plurality of videos watched by the user in history and the preset video set, the number of the videos is counted, and a recommendation parameter corresponding to each dimension feature information is calculated according to the preset weight parameter set corresponding to each dimension feature information and the counted number of the videos, so as to obtain a recommendation parameter set corresponding to the plurality of videos.
For another example, in an embodiment, the plurality of videos historically viewed by the user include some currently popular videos, which may include a first specific video, a second specific video, and so on, an image frame is extracted from the first specific video and the second specific video, an image frame set of the first specific video and the second specific video is obtained, and feature information of multiple dimensions of the target object, such as color feature information, style feature information, and so on, is extracted from the image frame set of the two videos.
The color feature information comprises first specific color feature information, second specific color feature information, third specific color feature information and the like, a preset specific weight parameter set corresponding to the color feature information is obtained from image frame sets of the two videos, and a recommended parameter corresponding to the color feature information can be calculated through the preset weight parameter corresponding to each obtained color feature information and the number of specific videos which are watched.
Similarly, recommendation parameters corresponding to feature information of other dimensions can also be calculated, and the recommendation parameters form a recommendation parameter set of each video.
104. And fusing the recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set, and sending recommendation contents to the terminal of the user according to the recommendation parameters in the target recommendation parameter set.
The fusion is to calculate recommendation parameters corresponding to the videos and a preset target weight value set to obtain a target recommendation parameter set. Establishing priorities for recommendation parameter sets corresponding to a plurality of videos, and calculating a target recommendation parameter set according to recommendation parameters corresponding to the plurality of videos and a preset target weight value set, namely, a process of fusing the recommendation parameter sets corresponding to the plurality of videos.
The target weight set can be set manually, and is adjusted and perfected continuously according to the feedback of the user at the later stage.
Optionally, in an embodiment, the content recommendation method may further include the following steps:
acquiring a second user set similar to the videos watched by the user history based on the videos watched by the user history;
acquiring a second watching video set of the second user set;
based on the plurality of videos and the second watching video set, calculating similar recommendation parameters corresponding to each dimension characteristic information to obtain similar recommendation parameter sets corresponding to the plurality of videos;
the sending of the recommended content to the terminal of the user according to the target recommendation parameter set includes:
fusing the target recommendation parameter set and the similar recommendation parameter set to obtain a fused target recommendation parameter set;
and sending the recommended content to the terminal of the user according to the fused target recommended parameter set.
And the historical watching video record of the second user set is similar to the historical watching video record of the user and is a user set with high similarity.
For example, based on the user history watching video record and the second watching video record set, acquiring the number of videos watched by the user, the number of videos watched by each second user and the number of videos watched by the user and the second user together; then according to the number of videos watched by the user, the number of videos watched by each second user and the number of videos watched by the user and the second user together, calculating the similarity of the videos watched by the user and each second user to obtain a set of similarity of the videos watched by the user and each second user, and then according to the set of similarity of the videos watched by the user to obtain a set of similarity weight parameters of the videos watched by the user and each second user; and calculating similar recommendation parameters corresponding to each dimension characteristic information of the user based on the similar weight parameter set of the watched video to obtain a similar recommendation parameter set, finally, fusing the target recommendation parameter set and the similar recommendation parameter set, and sending recommendation content to the terminal of the user according to the fused target recommendation parameter set.
As can be seen from the above, the original data of the content recommendation that needs to be acquired in this embodiment may be data in a video website, and the accuracy of the content recommendation may be improved.
The embodiments of the present application will be described from the perspective of a content recommendation device, which may be specifically integrated in a server.
As shown in fig. 3, a content recommendation method may specifically include the following steps:
201. the server acquires a plurality of videos historically watched by the user, and extracts image frames from the videos to obtain an image frame set of each video.
For example, image frames are extracted at fixed or varying time intervals from a plurality of videos that the user has watched before currently, and the image frames are collected into an image frame collection, wherein the plurality of videos may include different types of videos, such as comedy, literature, animation, and the like, and may also include some specific videos, such as popular videos currently liked by most users.
202. And the server extracts the multi-dimensional characteristic information of the target object from the image frame set of each video to obtain the multi-dimensional characteristic information of each video.
For example, multi-dimensional feature information of the target object is extracted from the image frame set, and when a person appears in the image frame set, the multi-dimensional feature information of the person, such as the style of clothes, the color of the clothes, and the like, is extracted.
The style of the clothes can be shirts, one-piece dresses, jeans and the like, and the color can be red, green, blue and the like.
Optionally, the extracted multi-dimensional feature information of the target object is not limited to the above-exemplified feature information, but may also include many other feature information, and the target object is not limited to a person, and may also be others.
203. And the server calculates the recommendation parameters corresponding to the dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos.
For example, in an embodiment, the plurality of videos historically viewed by the user include a first video and a second video, the feature information of multiple dimensions of the target object, such as color feature information of clothes, style feature information of clothes, and the like, is extracted from the image frame sets of the two videos, and in the image frame set, the color feature information may include red, green, blue, and the like, at this time, a recommended parameter of the color for the user, that is, a preference value representing a preference degree for the color, that is, a preference value corresponding to the color feature information, may be calculated, if in the image frame set, the number of frames of red in the image frame set of the first video is 3 frames, green is 7 frames, and blue is 10 frames, and the number of frames of red in the image frame set of the second video is 6 frames, green is 4 frames, and blue is 0 frame, as shown in the following table:
video Number of Red frames Number of frames of green Number of blue frames Total number of extracted frames
First video 3 7 10 20
Second video 6 4 0 10
Based on the above table, the preference values corresponding to the color feature information are respectively R1 Red color、R1 green colorAnd R1 blue colorWherein R is1 Red colorIndicating that in this embodiment, the user's preference for red is R1 Red colorIt can be understood that preference values, i.e. recommended parameters, corresponding to other dimension characteristic information are also named similarly, so that color characteristic informationThe preference value corresponding to the information is calculated according to the above table and the following calculation formula, which may be:
wherein R is1 color nRepresents R1 Red color、R1 green colorAnd R1 blue colorThe color feature information is equal to the recommended parameter value, and the number of color n frames represents, in this embodiment, the total number of frames of a certain color of the first video and the second video, for example, the number of red frames in the first video is 3, and the number of red frames in the second video is 6, then the total number of red frames is 9 frames, and the total extracted number of frames represents, in this embodiment, the sum of the number of red frames, the number of green frames, and the number of blue frames in the first video and the second video, and the specific algorithm is as follows:
R1 Red color：(3+6)×10/30＝3
R1 green color：(7+4)×10/30＝3.67
R1 blue color：(10+0)×10/30＝3.33
By mixing the above R1 Red color、R2 green color、R3 blue colorBy comparison of the calculated values of (A) to (B), knowing R2 green colorThe value of (3) is the maximum, it can be determined that the user prefers blue, and the preference value corresponding to one color feature information of the user is known as R2 green color3.67, that is, the preference value is the recommended parameter corresponding to the color feature information, where all the values are multiplied by one value 10 to facilitate statistics on the recommended parameter value, and not only are the values limited to 10, but also other values may be included.
Similarly, style feature information of the clothes of the target object is extracted from the image frame sets of the first video and the second video, and in the style feature information, the number of image frames of each style, for example, the number of image frames of a shirt, the number of image frames of a one-piece dress, and the number of image frames of jeans, is obtained, if the number of frames of the shirt in the image frame set of the first video is 2 frames, the number of frames of the one-piece dress is 8 frames, and the number of frames of the jeans is 0 frames, and the number of frames of the shirt in the image frame set of the second video is 3 frames, the number of frames of the one-piece dress is 1 frame, and the number of frames of the jeans is 6 frames, as shown in:
video Number of frames of shirt Frame number of one-piece dress Frame number of jeans Total number of extracted frames
First video 2 8 0 10
Second video 3 1 6 10
Based on the above table, the preference values corresponding to the style characteristic information are R respectively1 shirt、R1 one-piece dressAnd R1 JeansWherein R is1 shirtIn this embodiment, the user's preference value for shirts is R1 shirtIt can be understood that preference values, i.e., recommended parameters, corresponding to other dimension characteristic information are also named in the same way, and then the preference values corresponding to the style characteristic information can be calculated as:
wherein R is1 type nRepresents R1 shirt、R1 one-piece dressAnd R1 JeansThe recommended parameter value corresponding to the style characteristic information is equal to the style n frame number, which indicates the total frame number of a certain style of the first video and the second video in this embodiment, for example, the frame number of a shirt in the first video is 2, and the frame number of a shirt in the second video is 3, then the total frame number of the shirt is 5 frames, and the total extracted frame number indicates the sum of the frame number of the shirt in the first video and the second video, the frame number of the one-piece dress, and the frame number of jeans in this embodiment, and the specific algorithm is as follows:
R1 shirt：(2+3)×10/20＝2.5
R1 one-piece dress：(8+1)×10/20＝4.5
R1 Jeans：(0+6)×10/20＝3
By mixing the above R1 shirt、R2 one-piece dress、R3 jeansBy comparison of the calculated values of (A) to (B), knowing R2 one-piece dressThe value of the user is the maximum value, the fact that the user likes the one-piece dress relatively can be obtained, and the preference value corresponding to one style characteristic information of the user is known as R2 one-piece dressThe preference value is 4.5, that is, the preference value is the recommended parameter corresponding to the style characteristic information, and similarly, all the values are multiplied by one value 10 to facilitate statistics on the recommended parameter value, and not only are the value 10 limited, but also other values may be included.
Similarly, according to the method, the recommendation parameters of other dimensional multi-feature information are calculated, and the recommendation parameters are combined to form a recommendation parameter set.
For another example, in an embodiment, the videos historically viewed by the user include multiple types of videos, such as videos of types like comedy, literature, animation, and the like, feature information of multiple dimensions of the target object, such as color feature information of clothes, style feature information of clothes, and the like, is extracted from an image frame set of the video of the type, and a preset weight parameter set corresponding to the style feature information, that is, a weight value corresponding to each style feature information shown in the following table, is obtained from the image frame set of the videos of the type:
video Weight value of shirt Weight value of one-piece dress Weight value of jeans
Comedy type 5 3 8
Types of literature and art 1 9 5
Cartoon type 10 1 1
Wherein, a value of 5 in the above table indicates that in the video of the comedy type, the preset weight parameter value corresponding to the shirt in the style characteristic information is 5, and the same can be understood similarly to other values in the above table, at this time, if the user watches 10 videos, among which there are 2 videos of the comedy type, 6 videos of the literature type and 2 videos of the animation type, then based on the above table and the following formula, the preference values corresponding to the style characteristic information can be calculated as:
wherein R is1 type xRepresents R2 shirt、R2 one-piece dressAnd R2 JeansThe recommended parameter value corresponding to the style-waiting feature information, and AWeight value of style xFor obtaining the preset weighting parameter corresponding to the style characteristic information, for example, the preset weighting value 5 in the above table, etc. preset weighting parameter value, nType (B)The number of videos of a certain genre is represented, for example, the above-mentioned 2 videos of comedy genre, 6 videos of literature genre and 2 videos of animation genre, NTotal number of viewsIn this embodiment, the total number of videos watched by the user is expressed, that is, 10 videos are watched, and the specific algorithm is as follows:
R2 shirt：(5×2+1×6+10×2)/10＝3.5
R2 one-piece dress：(3×2+9×6+1×2)/20＝6.2
R2 Jeans：(8×2+5×6+1×2)/10＝4.8
By mixing the above R2 shirt、R2 one-piece dress、R2 JeansBy comparison of the calculated values of (A) to (B), knowing R2 one-piece dressThe value of the user is the maximum value, the fact that the user likes the one-piece dress relatively can be obtained, and the preference value corresponding to one style characteristic information of the user is known as R2 one-piece dressAnd 6.2, namely the preference value is the recommended parameter corresponding to the style characteristic information.
Similarly, according to the method, the recommendation parameters of other dimensional multi-feature information are calculated, and the recommendation parameters are combined to form a recommendation parameter set.
For another example, in an embodiment, the videos watched by the user include specific videos, such as popular videos, multi-dimensional feature information of the target object, such as feature information of color, style, and the like, is extracted from the image frame set obtained by obtaining the specific videos, and then a preset weight parameter set corresponding to the feature information is obtained, that is, a weight value corresponding to each color feature information is shown in the following table:
video Red weight value Green weight value Weight value of blue
First specific video 10 5 0
Second specific video 0 10 0
Wherein, a value 10 in the table indicates that in the first specific video, the preset weighting parameter value for red pair in the color feature information is 10, and the same can be understood in the same way as other values in the table, at this time, if a plurality of videos historically watched by the user includes exactly 2 videos of the first specific video and the second specific video, then based on the table, in this embodiment, the preference values corresponding to the color feature information may be calculated as:
wherein R is known3 color nRepresents R3 Red color、R3 green colorAnd R3 blue colorThe color feature information is equal to the corresponding recommended parameter value, and AColor n weight valueThe preset weight parameter value representing the color feature information of the first specific video and the second specific video in this embodiment, for example, the preset weight value such as 10 in the table above, and NSpecific video viewing numberIt is shown that in this embodiment, the plurality of videos viewed by the user includes NSpecific video viewing numberA specific video, NSpecific video viewing numberThe number of the specific videos watched by the user is shown, for example, it can be known from the above table that the user watches 2 specific videos of the first specific video and the second specific video, and the specific algorithm is as follows:
R3 Red color：(10+0)/2＝5
R3 green color：(10+5)/2＝7.5
R3 blue color：(0+0)/2＝0
By mixing the above R3 Red color、R3 green color、R3 blue colorBy comparison of the calculated values of (A) to (B), knowing R3 green colorThe value of the user is maximum, the user can be found to be more favorable for green, and the preference value corresponding to one style characteristic information of the user is known as R3 green color6.2, that is, the preference value is a recommended parameter corresponding to the style characteristic information, where the value 2 represents the number of specific videos viewed by the user.
Similarly, according to the method, the recommendation parameters of other dimensional multi-feature information are calculated, and the recommendation parameters are combined to form a recommendation parameter set.
204. And the server fuses the recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set, and sends recommendation content to the terminal of the user according to the recommendation parameters in the target recommendation parameter set.
For example, a priority is set for the obtained recommendation parameter set, a preset target weight value is obtained, a target recommendation parameter set is calculated according to the recommendation parameter sets and the preset target weight value, and then recommendation content is sent to a terminal of a user according to recommendation parameters in the target recommendation parameter set, wherein the recommendation content can be sent to the terminal of the user according to one recommendation parameter in the target recommendation parameter set, the recommendation content can also be sent to the terminal of the user according to a plurality of recommendation parameters in the target recommendation parameter set, and the target weight parameter value can be manually set, and is continuously adjusted and improved according to feedback of the user in a later period.
Optionally, in an embodiment, a second user group with the highest similarity to the historical viewing record of the user may also be obtained according to the historical viewing video record of the user, then according to the second user group, a similar recommendation parameter set of multiple videos of the user is calculated, then the similar recommendation parameter set and a target recommendation parameter set are fused to obtain a fused target recommendation parameter set, and finally, recommendation content is sent to the terminal of the user according to recommendation parameters in the fused target recommendation parameter set, for example, recommendation content may be sent to the terminal of the user according to one recommendation parameter in the fused target recommendation parameter set, or recommendation parameter content may be sent to the terminal of the user according to multiple recommendation parameters in the fused target recommendation parameter set.
For example, according to the historical watching video records of the user, a second user group with the highest similarity to the historical watching records of the user is obtained, then M historical watching video records of the user are obtained, wherein the value M is the number of videos historically watched by the user, then M historical watching video records are respectively obtained from each user of the second user group, if the number of watching videos of a part of users in the second user group is less than M, all the watching video records of the part of users who are less than M watching video records are obtained, for example, the number of the historical watching videos of the part of users is L (L < M).
Optionally, let K be the number of the historical videos that the user and each second user have viewed together, S be the similarity of the video that the user and each second user view, and the similarity S be:
according to the above formula, the similarity values corresponding to the user and the second user a, the second user B, the second user C, and the second user D in the second user group are calculated respectively, as shown in the following table:
second user A Second user B Second user C Second user D
Similarity to user 2/7 3/10 1/10 3/8
When the similarity value is determined, the corresponding similarity weight may be calculated according to the following formula, where the similarity weight value P is as follows:
wherein S isnRepresenting the similarity between the first user and a second user, such as the similarity between the user and the user A being 2/7, the similarity between the user and the user B being 3/10, etc., the similarity between the user and each second user can be calculated according to the above formulaThe similarity weight of the user, wherein the calculated similarity weight values are shown in the following table:
second user A Second user B Second user C Second user D
Similarity weight to user 80/297 84/297 28/297 105/297
Acquiring the recommendation parameters of the original feature information of each dimension of the second user group according to the historical recommendation record of the second user group, wherein the recommendation parameters are shown in the following table:
similar users Red colour Green colour Blue color
Second user A 5 3 8
Second user B 6 3 9
Second user C 4 9 3
Second user D 2 3 7
In this embodiment, the value 5 in the table indicates that the recommended parameter value corresponding to red in the color feature information of the user a is 5, similarly, the value 6 indicates that the recommended parameter value corresponding to red in the color feature information of the user B is 6, and so on.
The preference value corresponding to the color feature information can be calculated by the following formula:
R4 color n＝∑AUser x _ color n×PUser x similarity weight
Wherein R is known4 color nRepresents R4 Red color、R4 green colorAnd R4 blue colorThe color feature information is equal to the corresponding recommended parameter value, and AUser x _ color nA recommended parameter value corresponding to a color in the color feature information indicating a second user in this embodiment, such as a similarity weightA weight value, such as a recommended parameter value of 5 corresponding to red in the color feature information representing the user A, and so on, and PUser x similarity weightThen, a similarity weight value between the user and a second user is represented, for example, the similarity weight value between the user and the user a is 80/297, and the specific algorithm is as follows:
R4 Red color：5×80/297+6×84/297+4×28/297+2×105/297＝4.13
R4 green color：3×80/297+3×84/297+9×28/297+3×105/297＝3.57
R4 blue color：8×80/297+9×84/297+3×28/297+7×105/297＝7.46
By mixing the above R4 Red color、R4 green color、R4 blue colorBy comparison of the calculated values of (A) to (B), knowing R4 blue colorThe value of the user is maximum, the user can be found to be more favorable for green, and the preference value corresponding to one style characteristic information of the user is known as R4 blue colorAnd 7.46, namely the preference value is the similar recommended parameter corresponding to the style characteristic information.
Similarly, according to the method, similar recommendation parameters of other dimensional multi-feature information are calculated, the recommendation parameter sets are synthesized into a similar recommendation parameter set, then the similar recommendation parameter set and a target recommendation parameter set are fused to obtain fused target recommendation parameters, and finally recommendation content is sent to the video client of the user according to the recommendation parameters in the fused target recommendation parameter set.
As can be seen from the above, the original data of the content recommendation that needs to be acquired in this embodiment may be data in a video website, and the accuracy of the content recommendation may be improved.
In order to better implement the method, correspondingly, the embodiment of the application also provides a content recommendation device, wherein the content recommendation device can be specifically integrated in the server.
For example, as shown in fig. 4, the content recommendation apparatus may include a first acquisition unit 301, a first extraction unit 302, a second extraction unit 303, a first calculation unit 304, a fusion unit 305, and a transmission unit 306 as follows:
(1) a first acquisition unit 301;
a first obtaining unit 301, configured to obtain a plurality of videos historically viewed by a user.
(2) A first extraction unit 302;
a first extracting unit 302, configured to extract image frames from the video, and obtain an image frame set of each video.
(3) A second extraction unit 303;
the second extracting unit 303 is configured to extract multi-dimensional feature information of the target object from the image frame set of each video, so as to obtain the multi-dimensional feature information of each video.
(4) A first calculation unit 304;
the first calculating unit 304 is configured to calculate a recommendation parameter corresponding to each dimension feature information of each video, so as to obtain a recommendation parameter set corresponding to the multiple videos.
In one embodiment, the first computing unit 304 includes:
a first obtaining subunit 3041, configured to obtain, from the image frame set of each video, the number of target image frames corresponding to each dimension feature information;
the first calculating subunit 3042 is configured to calculate, based on the number of the target image frames and the number of the image frames in the image frame set, a recommended parameter corresponding to each piece of dimensional feature information, so as to obtain a recommended parameter set corresponding to the multiple videos.
In one embodiment, the first computing unit 304 includes:
a second obtaining subunit 3043, configured to obtain a preset weight parameter set corresponding to each dimension feature information;
the second calculating subunit 3044 is configured to calculate, according to the preset weight parameter set and the video attribute information corresponding to the multiple videos, a recommended parameter corresponding to each dimension feature information, so as to obtain a recommended parameter set corresponding to the multiple videos.
In an embodiment, the first computing unit 304 further includes:
a third obtaining subunit 3045, configured to obtain video types of the multiple videos, and count the number of each video type.
In one embodiment, the first calculating subunit 304 includes:
a third calculating subunit 3046, configured to calculate, according to the preset weight parameter set and the number of each video type, a recommended parameter corresponding to each dimensionality feature information, so as to obtain a recommended parameter set corresponding to the multiple videos.
In an embodiment, the first calculating subunit 304 further includes:
a selecting subunit 3047, configured to select a video that matches with a preset video from the multiple videos, obtain a matching video set, and count the number of videos in the matching video set.
In one embodiment, the first calculating subunit 304 includes:
a fourth calculating subunit 3048, configured to calculate, based on the preset weight parameter set and the number of videos in the matching video set, a recommended parameter corresponding to each dimensionality feature information, so as to obtain a recommended parameter set corresponding to the multiple videos.
(5) A fusion unit 305;
the fusion unit 305 is configured to fuse the recommendation parameter sets corresponding to the multiple videos to obtain a target recommendation parameter set.
(6) A transmitting unit 306;
a sending unit 306, configured to send recommended content to the terminal of the user according to the recommendation parameter in the target recommendation parameter set.
In an embodiment, the sending unit 306 includes:
a fusion subunit 3061, configured to fuse the target recommendation parameter set with the similar recommendation parameter set to obtain a fused target recommendation parameter set;
a sending subunit 3062, configured to send the recommended content to the terminal of the user according to the fused target recommendation parameter set.
In an embodiment, the content recommendation apparatus may further include:
a second obtaining unit 307, configured to obtain a second user set similar to the video watched by the user history based on the videos watched by the user history;
a third obtaining unit 308, configured to obtain a second viewing video set of the second user set;
the second calculating unit 309 is configured to calculate, based on the multiple videos and the second viewing video set, a similar recommendation parameter corresponding to each dimension feature information, so as to obtain a similar recommendation parameter set corresponding to the multiple videos.
In one embodiment, the second computing unit 309 includes:
a statistics subunit 3091, configured to count, based on the plurality of videos and the second viewing video set, the number of historical viewing videos corresponding to the user and the second users;
a fifth calculating subunit 3092, configured to calculate, according to the number of the historical viewing videos, similarities between the user and the viewing videos of the second users, so as to obtain a set of similarity between the viewing videos;
the sixth calculating subunit 3093 is configured to calculate, according to the set of similarity degrees of the videos to be watched, a similar recommendation parameter corresponding to each dimension feature information, so as to obtain a set of similar recommendation parameters corresponding to the multiple videos.
As can be seen from the above, in the content recommendation device of the present embodiment, the first acquisition unit 301 acquires a plurality of videos historically viewed by the user; then, the first extraction unit 302 extracts image frames from the video to obtain an image frame set of each video; extracting multi-dimensional feature information of the target object from the image frame set of each video by a second extraction unit 303 to obtain the multi-dimensional feature information of each video; the first calculating unit 304 calculates recommendation parameters corresponding to each dimension feature information of each video to obtain recommendation parameter sets corresponding to the multiple videos; the fusion unit 305 fuses the recommendation parameter sets corresponding to the multiple videos to obtain a target recommendation parameter set; and the sending unit 306 sends the recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set. The original data of the content recommendation, which needs to be acquired in the scheme, can be data in a video website, and the accuracy of the content recommendation can be improved.
Accordingly, an embodiment of the present application further provides a computer device, where the computer device may be a network-side device such as a server, and as shown in fig. 5, a schematic structural diagram of the server according to the embodiment of the present application is shown, specifically:
the computer device may include components such as a processor 401 of one or more processing cores, memory 402 of one or more computer-readable storage media, a power supply 403, and an input unit 404. Those skilled in the art will appreciate that the computer device configuration illustrated in FIG. 5 does not constitute a limitation of computer devices, and may include more or fewer components than those illustrated, or some components may be combined, or a different arrangement of components. Wherein:
the processor 401 is a control center of the computer device, connects various parts of the entire computer device using various interfaces and lines, and performs various functions of the computer device and processes data by running or executing software programs and/or modules stored in the memory 402 and calling data stored in the memory 402, thereby monitoring the computer device as a whole. Optionally, processor 401 may include one or more processing cores; preferably, the processor 401 may integrate an application processor, which mainly handles operating systems, user interfaces, application programs, etc., and a modem processor, which mainly handles wireless communications. It will be appreciated that the modem processor described above may not be integrated into the processor 401.
The memory 402 may be used to store software programs and modules, and the processor 401 executes various functional applications and data processing by operating the software programs and modules stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data created according to use of the computer device, and the like. Further, the memory 402 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device. Accordingly, the memory 402 may also include a memory controller to provide the processor 401 access to the memory 402.
The computer device further comprises a power supply 403 for supplying power to the various components, and preferably, the power supply 403 is logically connected to the processor 401 via a power management system, so that functions of managing charging, discharging, and power consumption are implemented via the power management system. The power supply 403 may also include any component of one or more dc or ac power sources, recharging systems, power failure detection circuitry, power converters or inverters, power status indicators, and the like.
The computer device may also include an input unit 404, the input unit 404 being operable to receive input numeric or character information and to generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.
Although not shown, the computer device may further include a display unit and the like, which are not described in detail herein. Specifically, in this embodiment, the processor 401 in the computer device loads the executable file corresponding to the process of one or more application programs into the memory 402 according to the following instructions, and the processor 401 runs the application programs stored in the memory 402, thereby implementing various functions as follows:
acquiring a plurality of videos historically watched by a user; extracting image frames from the video to obtain an image frame set of each video; extracting multi-dimensional characteristic information of a target object from an image frame set of each video to obtain the multi-dimensional characteristic information of each video; calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos; fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set; and sending the recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
For the above embodiments, reference may be made to the foregoing embodiments, and details are not described herein.
In one embodiment, the server may be a node in a distributed system, wherein the distributed system may be a blockchain system, and the blockchain system may be a distributed system formed by connecting a plurality of nodes through a network communication form. The nodes may form a Peer-To-Peer (P2P, Peer To Peer) network, and any type of computer device, such as a server, a terminal, and other electronic devices, may become a node in the blockchain system by joining the Peer-To-Peer network.
Therefore, the original data of the content recommendation, which needs to be acquired in the embodiment of the application, can be data in a video website, and the accuracy of the content recommendation can be improved.
It will be understood by those skilled in the art that all or part of the steps of the methods of the above embodiments may be performed by instructions or by associated hardware controlled by the instructions, which may be stored in a computer readable storage medium and loaded and executed by a processor.
To this end, embodiments of the present application provide a computer-readable storage medium, in which a plurality of instructions are stored, where the instructions can be loaded by a processor to execute the steps in any one of the content recommendation methods provided in the embodiments of the present application. For example, the instructions may perform the steps of:
acquiring a plurality of videos historically watched by a user; extracting image frames from the video to obtain an image frame set of each video; extracting multi-dimensional characteristic information of a target object from an image frame set of each video to obtain the multi-dimensional characteristic information of each video; calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos; fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set; and sending the recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
The above detailed implementation of each operation can refer to the foregoing embodiments, and is not described herein again.
Wherein the storage medium may include: read Only Memory (ROM), Random Access Memory (RAM), magnetic or optical disks, and the like.
Since the instructions stored in the storage medium may execute the steps in any content recommendation method provided in the embodiments of the present application, beneficial effects that can be achieved by any content recommendation method provided in the embodiments of the present application may be achieved, which are detailed in the foregoing embodiments and will not be described herein again.
The content recommendation method, device, terminal and computer-readable storage medium provided by the embodiments of the present application are described in detail above, and a specific example is applied in the present application to explain the principle and the implementation of the present application, and the description of the above embodiments is only used to help understand the method and the core idea of the present application; meanwhile, for those skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.

Claims (10)

1. A content recommendation method, comprising:
acquiring a plurality of videos historically watched by a user;
extracting image frames from the videos to obtain an image frame set of each video;
extracting multi-dimensional characteristic information of a target object from an image frame set of each video to obtain the multi-dimensional characteristic information of each video;
calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos;
fusing recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set;
and sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
2. The content recommendation method according to claim 1, wherein the calculating recommendation parameters corresponding to each dimension feature information of each video to obtain recommendation parameter sets corresponding to the plurality of videos comprises:
acquiring the number of target image frames corresponding to each dimension characteristic information from the image frame set of each video;
and calculating a recommendation parameter corresponding to each dimension characteristic information based on the number of the target image frames and the number of the image frames of the image frame set to obtain a recommendation parameter set corresponding to the plurality of videos.
3. The content recommendation method according to claim 1, wherein the calculating recommendation parameters corresponding to each dimension feature information of each video to obtain recommendation parameter sets corresponding to the plurality of videos includes:
acquiring a preset weight parameter set corresponding to each dimension characteristic information;
and calculating a recommendation parameter corresponding to each dimension characteristic information according to the preset weight parameter set and the video attribute information corresponding to the plurality of videos to obtain recommendation parameter sets corresponding to the plurality of videos.
4. The content recommendation method according to claim 3, characterized in that the method further comprises:
acquiring the video types of the plurality of videos, and counting the number of each video type;
the calculating recommendation parameters corresponding to each dimension feature information according to the preset weight parameter set and the video attribute information corresponding to the plurality of videos to obtain recommendation parameter sets corresponding to the plurality of videos includes:
and calculating recommendation parameters corresponding to each dimension characteristic information according to the preset weight parameter set and the number of each video type to obtain recommendation parameter sets corresponding to the plurality of videos.
5. The content recommendation method according to claim 3, characterized in that the method further comprises:
selecting videos matched with a preset video from the videos to obtain a matched video set, and counting the number of the videos of the matched video set;
the calculating recommendation parameters corresponding to each dimension feature information according to the preset weight parameter set and the video attribute information corresponding to the plurality of videos to obtain recommendation parameter sets corresponding to the plurality of videos includes:
and calculating recommendation parameters corresponding to each dimension characteristic information based on the preset weight parameter set and the number of videos in the matched video set to obtain recommendation parameter sets corresponding to the videos.
6. The content recommendation method according to claim 1, further comprising:
acquiring a second user set similar to the videos watched by the user history based on the videos watched by the user history;
acquiring a second watching video set of the second user set;
based on the plurality of videos and the second watching video set, calculating similar recommendation parameters corresponding to each dimension characteristic information to obtain similar recommendation parameter sets corresponding to the plurality of videos;
the sending of the recommended content to the terminal of the user according to the target recommendation parameter set includes:
fusing the target recommendation parameter set and the similar recommendation parameter set to obtain a fused target recommendation parameter set;
and sending recommendation content to the terminal of the user according to the fused target recommendation parameter set.
7. The content recommendation method according to claim 6, wherein the calculating similar recommendation parameters corresponding to each dimension feature information based on the plurality of videos and the second viewing video set to obtain similar recommendation parameter sets corresponding to the plurality of videos comprises:
counting the number of the historical watching videos corresponding to the users and the second users on the basis of the plurality of videos and the second watching video set;
according to the number of the historical watching videos, calculating the similarity of the user and the videos watched by the second users to obtain a watching video similarity set;
and calculating similar recommendation parameters corresponding to each dimension characteristic information according to the watching video similarity set to obtain similar recommendation parameter sets corresponding to the videos.
8. A content recommendation apparatus characterized by comprising:
the device comprises an acquisition unit, a display unit and a display unit, wherein the acquisition unit is used for acquiring a plurality of videos historically watched by a user;
the first extraction unit is used for extracting image frames from the videos to obtain an image frame set of each video;
the second extraction unit is used for extracting multi-dimensional characteristic information of the target object from the image frame set of each video to obtain the multi-dimensional characteristic information of each video;
the calculation unit is used for calculating recommendation parameters corresponding to each dimension characteristic information of each video to obtain recommendation parameter sets corresponding to the videos;
the fusion unit is used for fusing the recommendation parameter sets corresponding to the videos to obtain a target recommendation parameter set;
and the sending unit is used for sending recommended content to the terminal of the user according to the recommended parameters in the target recommended parameter set.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps in the content recommendation method according to any one of claims 1 to 7 when executing the program.
10. A computer-readable storage medium storing a plurality of instructions adapted to be loaded by a processor to perform the steps of the content recommendation method of any one of claims 1 to 7.
CN201910838291.4A 2019-09-05 2019-09-05 Content recommendation method and device, computer equipment and computer-readable storage medium Pending CN110598618A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910838291.4A CN110598618A (en) 2019-09-05 2019-09-05 Content recommendation method and device, computer equipment and computer-readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910838291.4A CN110598618A (en) 2019-09-05 2019-09-05 Content recommendation method and device, computer equipment and computer-readable storage medium

Publications (1)

Publication Number Publication Date
CN110598618A true CN110598618A (en) 2019-12-20

Family

ID=68857758

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910838291.4A Pending CN110598618A (en) 2019-09-05 2019-09-05 Content recommendation method and device, computer equipment and computer-readable storage medium

Country Status (1)

Country Link
CN (1) CN110598618A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112100504A (en) * 2020-11-03 2020-12-18 北京达佳互联信息技术有限公司 Content recommendation method and device, electronic equipment and storage medium
CN112351345A (en) * 2020-11-04 2021-02-09 深圳Tcl新技术有限公司 Control method and device of recommended content, smart television and storage medium
CN112749297A (en) * 2020-03-03 2021-05-04 腾讯科技（深圳）有限公司 Video recommendation method and device, computer equipment and computer-readable storage medium
CN113395550A (en) * 2020-03-13 2021-09-14 浙江大搜车软件技术有限公司 Video recommendation method, device and system and electronic equipment
CN113610557A (en) * 2021-07-09 2021-11-05 苏州众言网络科技股份有限公司 Advertisement video processing method and device
CN114449342A (en) * 2022-01-21 2022-05-06 腾讯科技（深圳）有限公司 Video recommendation method and device, computer readable storage medium and computer equipment
CN114915816A (en) * 2021-12-30 2022-08-16 天翼数字生活科技有限公司 User watching behavior acquisition and release method and system

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108228882A (en) * 2018-01-26 2018-06-29 维沃移动通信有限公司 The recommendation method and terminal device of a kind of audition for the songs segment
CN108471544A (en) * 2018-03-28 2018-08-31 北京奇艺世纪科技有限公司 A kind of structure video user portrait method and device
CN110008376A (en) * 2019-03-22 2019-07-12 广州新视展投资咨询有限公司 User's portrait vector generation method and device
CN110020111A (en) * 2017-09-21 2019-07-16 腾讯科技（深圳）有限公司 Stroke recommended method, device, computer equipment and storage medium

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110020111A (en) * 2017-09-21 2019-07-16 腾讯科技（深圳）有限公司 Stroke recommended method, device, computer equipment and storage medium
CN108228882A (en) * 2018-01-26 2018-06-29 维沃移动通信有限公司 The recommendation method and terminal device of a kind of audition for the songs segment
CN108471544A (en) * 2018-03-28 2018-08-31 北京奇艺世纪科技有限公司 A kind of structure video user portrait method and device
CN110008376A (en) * 2019-03-22 2019-07-12 广州新视展投资咨询有限公司 User's portrait vector generation method and device

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112749297A (en) * 2020-03-03 2021-05-04 腾讯科技（深圳）有限公司 Video recommendation method and device, computer equipment and computer-readable storage medium
CN112749297B (en) * 2020-03-03 2023-07-21 腾讯科技（深圳）有限公司 Video recommendation method, device, computer equipment and computer readable storage medium
CN113395550A (en) * 2020-03-13 2021-09-14 浙江大搜车软件技术有限公司 Video recommendation method, device and system and electronic equipment
CN112100504A (en) * 2020-11-03 2020-12-18 北京达佳互联信息技术有限公司 Content recommendation method and device, electronic equipment and storage medium
CN112100504B (en) * 2020-11-03 2021-09-10 北京达佳互联信息技术有限公司 Content recommendation method and device, electronic equipment and storage medium
WO2022095585A1 (en) * 2020-11-03 2022-05-12 北京达佳互联信息技术有限公司 Content recommendation method and device
CN112351345A (en) * 2020-11-04 2021-02-09 深圳Tcl新技术有限公司 Control method and device of recommended content, smart television and storage medium
CN113610557A (en) * 2021-07-09 2021-11-05 苏州众言网络科技股份有限公司 Advertisement video processing method and device
CN114915816A (en) * 2021-12-30 2022-08-16 天翼数字生活科技有限公司 User watching behavior acquisition and release method and system
CN114449342A (en) * 2022-01-21 2022-05-06 腾讯科技（深圳）有限公司 Video recommendation method and device, computer readable storage medium and computer equipment
CN114449342B (en) * 2022-01-21 2024-02-27 腾讯科技（深圳）有限公司 Video recommendation method, device, computer readable storage medium and computer equipment

Similar Documents

Publication Publication Date Title
CN110598618A (en) Content recommendation method and device, computer equipment and computer-readable storage medium
CN108829764B (en) Recommendation information acquisition method, device, system, server and storage medium
WO2020048084A1 (en) Resource recommendation method and apparatus, computer device, and computer-readable storage medium
CN104615775B (en) A kind of user&#39;s recommendation method and device
CN110601956B (en) Content aggregation method and device, computer equipment and storage medium
JP5443854B2 (en) Computer-implemented method to facilitate social networking based on fashion-related information
CN106202316A (en) Merchandise news acquisition methods based on video and device
CN107894998B (en) Video recommendation method and device
CN106202317A (en) Method of Commodity Recommendation based on video and device
TW201524567A (en) Recommendation method, device, and server
US20180096410A1 (en) Method and Apparatus for Providing Matching Information of Business Object
JP5442547B2 (en) Content recommendation apparatus and method
US20160210367A1 (en) Transition event detection
CN105100164B (en) Network service recommends method and apparatus
CN108900924A (en) The method and apparatus of commending friends in direct broadcasting room
CN108156528A (en) A kind of method for processing video frequency and device
JP5999425B2 (en) Information processing apparatus and information processing method
CN108122126B (en) Method for extending association rule, device using same and computer readable medium
CN110110206B (en) Method, device, computing equipment and storage medium for mining and recommending relationships among articles
JP2016071881A (en) Item recommendation system
CN105893515B (en) Information processing method and server
CN113225580A (en) Live broadcast data processing method and device, electronic equipment and medium
WO2016188278A1 (en) Method and device for collecting information about service object
WO2013018913A1 (en) Video processing system, method of determining viewer preference, video processing apparatus, and control method and control program therefor
CN111447260A (en) Information pushing and information publishing method and device

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20191220