CN110941738B - Recommendation method and device, electronic equipment and computer-readable storage medium - Google Patents
Recommendation method and device, electronic equipment and computer-readable storage medium Download PDFInfo
- Publication number
- CN110941738B CN110941738B CN201911184565.9A CN201911184565A CN110941738B CN 110941738 B CN110941738 B CN 110941738B CN 201911184565 A CN201911184565 A CN 201911184565A CN 110941738 B CN110941738 B CN 110941738B
- Authority
- CN
- China
- Prior art keywords
- preset
- crowd
- user
- behavior data
- target
- 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.)
- Active
Links
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
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/735—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/7867—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title and artist information, manually generated time, location and usage information, user ratings
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Library & Information Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention provides a recommendation method and device, electronic equipment and a computer readable storage medium, and belongs to the technical field of computers. According to the method, when it is detected that a user to be identified watches a video, historical behavior data generated when the user to be identified watches the video is acquired, the crowd category to which the user to be identified belongs is determined according to the historical behavior data, then, whether the crowd category to which the user to be identified belongs includes at least two crowd categories is judged, if the crowd category to which the user to be identified belongs includes at least two crowd categories, current behavior data generated when the user to be identified watches the video at present is acquired, then, a target crowd category is determined according to the current behavior data, and finally, target push content corresponding to the target crowd category is pushed to the user to be identified. In this way, the recommended content can be ensured to be the content which is interesting to the user to be identified who is watching the video at present, and the recommendation effect can be further ensured.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a recommendation method and apparatus, an electronic device, and a computer-readable storage medium.
Background
In order to improve the experience of the user in watching the video, it is often necessary to recommend a suitable video to the user when the user watches the video using the target application, so that the user can select to watch the video.
In the prior art, a user portrait is generally constructed according to user behavior data, and then corresponding content is recommended to a device used by a user based on the user portrait. However, in practical applications, a situation may occur in which the device is used by a plurality of users in a mixed manner. Therefore, once the device is used by people of other types, the original user portrait is interfered, and further contents which are not interesting to the original user are recommended to the original user, so that the recommendation effect is poor.
Disclosure of Invention
The invention provides a recommendation method, a recommendation device, electronic equipment and a computer-readable storage medium, which are used for solving the problem of poor recommendation effect.
In a first aspect of the present invention, there is provided a recommendation method, including:
when a user to be identified watches a video, acquiring historical behavior data generated when the user to be identified watches the video in the past;
determining the crowd category to which the user to be identified belongs according to the historical behavior data;
judging whether the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
if the crowd categories to which the user to be identified belongs comprise at least two crowd categories, acquiring current behavior data generated when the user to be identified watches videos currently;
determining the target crowd category according to the current behavior data;
and pushing target push contents corresponding to the target crowd category to the user to be identified.
In a second aspect of the present invention, there is provided a recommendation apparatus comprising:
the device comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring historical behavior data generated when a user to be identified watches a video in the past;
the first determining module is used for determining the crowd category to which the user to be identified belongs according to the historical behavior data;
the judging module is used for judging whether the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
the second obtaining module is used for obtaining current behavior data generated when the user to be identified watches the video currently if the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
the second determining module is used for determining the target crowd type according to the current behavior data;
and the pushing module is used for pushing target pushing content corresponding to the target crowd category to the user to be identified.
In yet another aspect of the present invention, there is also provided a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to execute any of the recommendation methods described above.
In yet another aspect of the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform any of the recommendation methods described above.
The recommendation method provided by the embodiment of the invention can acquire historical behavior data generated when a user to be identified watches a video, and determines the crowd category to which the user to be identified belongs according to the historical behavior data, then judges whether the crowd category to which the user to be identified belongs includes at least two crowd categories, and acquires current behavior data generated when the user to be identified watches the video, then determines the target crowd category according to the current behavior data, and finally pushes target push content corresponding to the target crowd category to the user to be identified if the crowd category to which the user to be identified belongs includes at least two crowd categories. Therefore, under the condition that the portrait of the user is damaged due to mixed use of different crowds in the process of watching the video in the past, the recommended content can be ensured to be the content which is interesting to the user to be identified and watching the video at present by a recommending mode based on the target crowd category, and the recommending effect can be further ensured.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below.
FIG. 1 is a flow chart illustrating the steps of a recommendation method according to an embodiment of the present invention;
FIG. 2-1 is a flow chart illustrating steps of another recommendation method provided by an embodiment of the present invention;
FIG. 2-2 is a system diagram according to an embodiment of the present invention;
FIG. 3 is a block diagram of a recommendation device provided by an embodiment of the present invention;
fig. 4 is a structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although illustrative embodiments of the invention are shown in the drawings.
Fig. 1 is a flowchart illustrating steps of a recommendation method according to an embodiment of the present invention, and as shown in fig. 1, the method may include:
In the embodiment of the present invention, when a play operation for a certain video is detected, or when it is detected that a device used by a user to be identified is in a process of playing a video, it is determined that the user to be identified watches the video. The historical behavior data may be data that reflects the interaction behavior with the target application during previous viewing of the video. The target application is an application program used when the identified user watches the video.
And step 102, determining the crowd category to which the user to be identified belongs according to the historical behavior data.
In the embodiment of the invention, the crowd types can be preset according to the attribute characteristics of the characters, and different people can belong to different crowd types. For example, the demographic categories may include adults, children, or men, women, or the elderly, middle aged, young, and so on. The content of interest of people belonging to the same crowd category tends to have similarity, while the content of interest of people belonging to different crowd categories tends to be different, so that the interaction behavior with the target application is different when different crowds watch videos. Since the historical behavior data can reflect the interaction behavior with the target application during the past video watching process. Therefore, in this step, the category of the crowd to which the user to be identified who has watched the video in the past belongs can be inferred by combining the historical behavior data.
And 103, judging whether the crowd categories to which the user to be identified belongs comprise at least two crowd categories.
In the process that a user watches videos in the past, the situation that multiple persons use the equipment in a mixed mode may exist, the situation that the multiple persons use the equipment in a mixed mode may also not exist, and under the situation that the multiple persons use the equipment in a mixed mode, the types of the groups of the persons determined in the steps can be multiple. Therefore, in this step, it may be determined whether the crowd categories to which the user to be identified belongs include at least two crowd categories. Specifically, the number of the crowd categories determined in the foregoing steps may be detected, where the number is an integer. Then, it is determined whether the number is greater than 1. If the number is greater than 1, at least two categories of people may be considered to be included, and if not greater than 1, at least two categories of people may be considered to be excluded.
And step 104, if the crowd categories to which the user to be identified belongs comprise at least two crowd categories, acquiring current behavior data generated when the user to be identified watches the video currently.
In the embodiment of the invention, if the crowd categories to which the user to be identified belongs include at least two crowd categories, it can be considered that the situation that different crowds use the equipment for watching in a mixed manner exists in the process of watching the video in the past, and the user portrait for video recommendation is interfered. In this case, if the user image before the user is used directly for recommendation, the recommended video may not be suitable for the current user, and the recommendation effect may be poor. Therefore, in the step, the current behavior data generated when the user to be identified watches the video currently can be further obtained, so that in the subsequent step, more accurate recommendation can be performed based on the current behavior data.
And 105, determining the target crowd category according to the current behavior data.
The current behavior data may be determined according to a real-time log generated by the target application, and the current behavior data may reflect an interaction behavior performed between the user to be identified and the target application in a current video watching process. And determining the target crowd category to which the user to be identified who is watching the video currently belongs according to the current behavior data, wherein the obtained target crowd category can accurately reflect the crowd attribute to which the user to be identified who is watching the video currently belongs.
And 106, pushing target push contents corresponding to the target crowd category to the user to be identified.
In this step, the target push content corresponding to the target crowd category may be a content that is interested by the crowd of the target crowd category. Because the user to be identified who is watching the video at present belongs to the target crowd category, when pushing is carried out according to the target crowd category, the pushed target push content can be ensured to a certain extent, and is the content which is interesting to the user to be identified who is watching the video at present, so that the recommendation effect is ensured.
In summary, in the recommendation method provided in the embodiment of the present invention, when it is detected that a user to be identified watches a video, historical behavior data generated when the user to be identified watches the video in the past is obtained, a crowd category to which the user to be identified belongs is determined according to the historical behavior data, then, it is determined whether the crowd category to which the user to be identified belongs includes at least two crowd categories, if the crowd category to which the user to be identified belongs includes at least two crowd categories, current behavior data generated when the user to be identified watches the video at present is obtained, then, a target crowd category is determined according to the current behavior data, and finally, a target push content corresponding to the target crowd category is pushed to the user to be identified. Therefore, under the condition that the portrait of the user is damaged due to mixed use of different crowds in the process of watching the video in the past, the recommended content can be ensured to be the content which is interesting to the user to be identified and watching the video at present by a recommending mode based on the target crowd category, and the recommending effect can be further ensured.
Fig. 2-1 is a flowchart illustrating steps of another recommended method according to an embodiment of the present invention, and as shown in fig. 2-1, the method may include:
Specifically, in this step, the acquisition of the historical behavior data may be realized by the following substeps (1) to (3):
substep (1): acquiring a history log generated by the application program within a preset time length from a history log file corresponding to a target application; the target application is an application program used by the user to be identified when watching the video.
The log, also called log, is usually a record of completed processing by a system or an application, and each log carries a generation time, which has no fixed format, is usually a text file, and may be in other formats. In this step, the history log file may be generated by an interactive behavior with a target application in a process of viewing a video in the past, and a log for recording each operation of a user is correspondingly generated for each operation of the user. For example, clicking and playing a video on a target application will generate a corresponding log. Further, the generated logs are often sent to a background server specially used for storing log files, where sending of the logs may be implemented by an automatic reference notification (ping back) technique, and the background server may be the same server as the server performing this step or a different server, which is not limited in this embodiment of the present invention. Accordingly, according to the generation time of each log, the history logs generated within the preset time length can be searched from the history log files. The preset time period may be preset according to actual conditions, for example, the preset time period may be within one day, within three days, or within 1 week, and the like, which is not limited in this embodiment of the present invention.
Substep (2): and for each history log, extracting information of a preset type from the history log.
Specifically, each history log may be detected to determine the preset type of information in the history log, and then, the information may be extracted. The preset type of information may be: the device identification of the user, the video identification, the object of the user behavior and the specific operation parameters of the user to the behavior object. The device identification of a user can be represented by a user _ id field, the device identification of a user can be used for uniquely representing a user, the video identification can be represented by a tvid field, a video identification can be used for uniquely representing a video, the object of user behavior can be represented by an action _ type field, the object of user behavior represents a specific operation performed by the user, and can include specific entities such as "click", "watch", "comment", and the like, and the specific operation parameter of the user on the behavior object can be represented by an < action details > field, and can represent the specific operation parameter of the user on the behavior object, for example, it can be the "watching duration" of the video watched by the user, the "comment content" of the video commented by the user, and the like.
The substep (3) combines the information of the preset types according to a preset information format to obtain the historical behavior data; the historical behavior data is used for representing specific information of various behaviors performed on the video by the historical user.
Specifically, the preset information format may be represented as a quadruple format, where the quadruple may be represented as: (user _ id, tvid, action _ type, < action details >) which can represent the specific information of the user's actions in the history log, and correspondingly, these information can be combined according to the quadruple format to obtain a quadruple information, wherein a quadruple information is a piece of history action data.
Because the log is often a text type file, the aggregation of the historical log is realized by extracting the preset type of information and combining the information according to the preset information format in the step, and the historical log is converted into structured data, and the structured data is more convenient to process, so that the processing efficiency of the data in the subsequent steps can be improved. It should be noted that the format of the preset information and the preset type for extracting information may also be adjusted according to actual requirements, which is not limited in the embodiment of the present invention. Since the number of history logs may be large, in a specific implementation, these sub-steps of extracting the historical behavior data from the history logs may be implemented by using a general-purpose computing engine (spark), which is a fast general-purpose computing engine designed for large-scale data processing. In this way, it is ensured that the operation of extracting the historical behavior data can be smoothly completed.
Specifically, this step can be realized by the following substeps (4) to (6):
substep (4): acquiring first preset judgment conditions corresponding to various preset crowd categories.
Due to the fact that videos of different people are different in interest, behavior characteristics of different people are different when the target application is used, for example, adult users cannot click a large amount of children videos in the using process. Therefore, in the embodiment of the present invention, according to behavior characteristics of different people when using a target application, corresponding first preset judgment conditions may be set for different people categories, where the first preset judgment conditions may represent specific information that needs to be satisfied by behaviors that are executed on a video by historical users, that is, users who watch the video in the past. For example, the first preset determining condition may be a first preset determining condition corresponding to a preset crowd category "pupil", a first preset determining condition corresponding to a preset crowd category "junior school" and a first preset determining condition corresponding to a preset crowd category "senior school student", and accordingly, the 2 first preset determining conditions may be obtained, so as to obtain the first preset determining conditions corresponding to multiple preset crowd categories. Further, the first preset judgment condition may also include a first preset judgment condition corresponding to a preset crowd category "child" and a first preset judgment condition corresponding to a preset crowd category "adult", and accordingly, the two first preset judgment conditions may be obtained, so as to obtain the first preset judgment conditions corresponding to multiple preset crowd categories. Specifically, the first preset judgment conditions may be stored in a specific storage location in advance, and when the crowd category needs to be determined according to the historical behavior data, the first preset judgments may be read from the specific storage location. The specific storage location may be selected according to actual situations, which is not limited in this embodiment of the present invention.
Substep (5): and respectively matching the historical behavior data by utilizing each first preset judgment condition.
Further, the behavior characteristics represented by the historical behavior data may be used to match with a first preset judgment condition corresponding to each preset crowd category, and specifically, if the historical behavior data can meet specific information that needs to be met by the behavior executed by the historical user on the video, which is represented by the first preset judgment condition, the historical behavior data and the first preset judgment condition are considered to be matched. For example, assume that the first preset determination condition is: the last 8 displayed non-child videos were no-click, and 80% of the last 12 clicks were child videos. The historical behavior data indicates that 90% of the videos clicked by the historical users last 12 are children videos, and non-children videos are not clicked, so that the two videos can be considered to be matched.
Substep (6): and determining a preset crowd category corresponding to a first preset judgment condition matched with the historical behavior data as the belonged crowd category.
For example, assuming that the historical behavior data matches a first preset judgment condition corresponding to a preset crowd category "child", the preset crowd category "child" may be determined as the crowd category to which the user to be identified belongs. It should be noted that, in the embodiment of the present invention, time and/or location of generation of a history log corresponding to each piece of historical behavior data may also be used as historical scene information corresponding to the historical behavior data; then, determining a scene type to which each piece of historical behavior data belongs based on historical scene information and preset scene information corresponding to a plurality of preset scene types; and then, for each scene category, matching historical behavior data belonging to the scene category according to preset judgment conditions corresponding to different crowd categories under the scene category to determine the crowd category to which the scene category belongs. The embodiment of the present invention is not limited thereto. The judgment accuracy can be improved by performing targeted judgment on the scenes, but the time span of the historical behavior data is often large, so that the historical behavior data can be divided into a plurality of scenes by performing judgment after dividing different scenes, and the workload of judgment operation is increased. Therefore, in the embodiment of the invention, the same standard is adopted for judging all the historical behavior data, and the processing resources required by the judgment operation can be reduced to a certain extent.
Specifically, the specific implementation manner of this step may refer to step 103, which is not described herein again in this embodiment of the present invention.
If the crowd category to which the user to be identified belongs does not include at least two crowd categories, step 204 is executed. Otherwise, step 205 is performed.
And 204, taking the crowd category to which the user to be identified belongs as the target crowd category.
If the crowd categories to which the user to be identified belongs do not include at least two crowd categories, it can be considered that the situation that different crowds use the equipment in a mixed manner for watching does not exist in the process of watching the video in the past, and the user portrait for video recommendation is not interfered. At this time, the crowd category to which the user to be identified belongs may be taken as the target crowd category, so that recommendation is performed based on the crowd category in subsequent steps. The crowd category to which the user to be identified belongs is determined based on historical behavior data, and the time span of the historical behavior data is often larger than that of the current behavior data, so that the data volume of the historical behavior data is often larger than that of the current behavior data. Therefore, under the condition that the portrait of the user is not interfered, the crowd category determined based on the historical behavior data is used as the target crowd category, so that the target crowd category is more accurate, the target crowd category is determined in time, and the subsequent recommendation effect can be improved. Meanwhile, the analysis and determination are not needed to be further carried out based on the real-time log, so that the processing amount can be reduced to a certain extent, and the processing resources are saved.
And step 205, acquiring current behavior data generated when the user to be identified watches the video currently.
In this step, the current behavior data may be determined according to a real-time log generated by the target application, the real-time log may be generated in real time during an interaction process between the current user, that is, a user to be identified who currently watches the video, and the target application, and the current behavior data may represent specific information of each behavior that the current user performs on the video at the current time. For example, a real-time log may be obtained, the real-time log is used as a real-time stream in a kafka (kafka) open source stream processing platform, and the real-time stream in the platform is processed in a streaming (streaming) task stream manner, so as to extract current behavior data from the real-time log. Specifically, the implementation manner of extracting the current behavior data from the real-time log may be described in association with the parameters in the foregoing steps, and details of the embodiment of the present invention are not described herein.
And step 206, determining the target crowd type according to the current behavior data.
Specifically, this step can be realized by the following substeps (7) to (10):
substep (7): determining a scene type to which the current scene belongs based on the current scene information and preset scene information corresponding to a plurality of preset scene types; the current scene information includes at least a current time and/or a current location.
In this step, the current scene information may be obtained first, for example, in an example where the server executes the recommendation method, the server may send a time obtaining instruction to the terminal to determine the current time, or query the current time from the network, and so on. Then, a location acquisition instruction is sent to the terminal to determine the current location. Of course, the time and place acquisition instruction may be sent, and the current time and place may be acquired at the same time. Or only one of the scene information is acquired as the current scene information, which is not limited in the embodiment of the present invention. The terminal may be a device used by the user to be identified when watching the video.
Further, the current scene information may be matched with preset scene information corresponding to a plurality of preset scene categories to determine a scene category to which the current scene belongs. The preset scene information corresponding to different scene categories can be preset according to actual requirements, and the preset scene information is not limited in the embodiment of the invention. For example, in the case that the scene information only includes time, the preset scene information of the scene type I may be between 16 and 22 on weekend holidays, legal holidays and non-holidays, and the preset scene information of the scene type II may be between 22 and 16 on the next day. Assuming that the current scene information is 17 points, since 17 points satisfy the preset scene information of the scene type I, it can be determined that the scene type to which the current scene belongs is the scene type I.
Substep (8): and acquiring second preset judgment conditions corresponding to various preset crowd categories under the scene category.
Because the real-time log is generated at the current time, the time span of the current behavior data is often very small, and the behavior characteristics of different crowds may be different in different scenes, in the embodiment of the invention, different second preset judgment conditions are set for different preset crowd categories in different scene categories, and the second preset judgment conditions represent specific information required to be met by behaviors executed by the current user on the video. For example, taking the preset crowd category as the child, the second preset determination condition corresponding to the child category in the scene category I may be: "the non-child videos displayed in the last 7 times have no click, and 70% of the non-child videos shot in the last 12 times are child videos, and the last 2 times of click videos are child videos", the second preset judgment condition corresponding to the child category under the scene category I may be: "the last 12 non-child videos shown are no-click, and 90% of the last 12 clicks are child videos, and the top 4 clicks are all child videos". Therefore, different conditions under different scenes are adopted for judgment, and the scene can be pertinently judged under the condition of not consuming too much resources, so that the judgment accuracy is improved to a certain degree. Specifically, the manner of obtaining the second preset judgment condition may refer to the manner of obtaining the first preset judgment condition, and the second preset judgment condition may be stored in a specific storage location in advance, and when the category of the target group needs to be determined according to the current behavior data, the second preset judgment may be read from the specific storage location. The specific storage location may be selected according to actual situations, which is not limited in this embodiment of the present invention.
Substep (9): and respectively matching the current behavior data by utilizing each second preset judgment condition.
Further, the behavior characteristics represented by the current behavior data may be used to match with each second preset judgment condition, specifically, if the current behavior data can satisfy the specific information that the behavior executed by the current user on the video, which is represented by the second preset judgment condition, needs to be satisfied, the current behavior data and the second preset judgment condition are considered to be matched.
And (10) determining a preset crowd category corresponding to a second preset judgment condition matched with the current behavior data as the target crowd category.
For example, if the crowd category corresponding to the second preset judgment condition matched with the current behavior data is a child, in this step, the preset crowd category "child" may be determined as the target crowd category.
Further, in order to improve the accuracy of the determination, in the embodiment of the present invention, before the sub-step (10), the current behavior data may be verified based on a preset verification condition, and if the current behavior data passes the verification, the step is executed. The preset verification condition may be a condition stricter than any second preset judgment condition, and may represent specific information that needs to be satisfied by a current behavior executed by the user on the video, for example, the preset verification condition may be: "the last 14 non-child videos shown are non-click, and 95% of the last 12 click videos are child videos, and the last 6 click videos (including this time) are all child videos". In this step, the current behavior data may be matched according to the verification condition, and if the current behavior data is matched with the verification condition, the verification may be considered to be passed. In this way, the reliability of the determination result can be improved by performing the verification using a more strict condition. Specifically, the matching manner may refer to the foregoing steps, and details are not described herein in this embodiment of the present invention. Of course, if the verification fails, the step of respectively matching the current behavior data by using each second preset judgment condition may be executed again, and the step may be executed after the step is executed again.
It should be noted that, in the embodiment of the present invention, when a new real-time log is generated by a target application, or when a duration between a current time and a target time meets a preset duration threshold, the step of determining the target crowd category through the current behavior data may be re-executed, so as to update the target crowd category. The target time is the time for determining the category of the target crowd, the real-time log is generated when the target application interacts with the user, and the preset time length threshold value can be set according to actual requirements. Specifically, the foregoing steps may be re-executed to re-determine the target group category, and then the re-determined target group category is used to cover the previously stored target group category, thereby implementing the update. Therefore, the target crowd category can be ensured to be accurate by updating when a new real-time log exists or the target crowd category is determined for a long time before.
Meanwhile, because the current behavior data is determined according to the real-time log generated by the target application, and the real-time log is generated according to the operation of the user on the target application, when the target crowd category is determined based on the current behavior data, the user needs to wait for a certain amount of operation to be executed, and thus, a certain delay is caused to the whole recommendation process. In the embodiment of the present invention, the historical behavior data is obtained first, the crowd category to which the user to be identified belongs is determined based on the historical behavior data, and the operation of determining the target crowd category based on the current behavior data is executed only when the crowd category to which the user to be identified belongs includes at least two crowd categories. Thus, the number of times the step is performed can be reduced, thereby reducing the delay to some extent.
And step 207, pushing target push contents corresponding to the target crowd category to the user to be identified.
Taking the steps executed by the server, and the user to be identified watches the video through the terminal capable of interacting with the server as an example, in the step, the target crowd category can be stored to a preset storage position, so that the terminal can read the target crowd category from the preset storage position, and push the target push content for the user to be identified based on the target recommendation strategy corresponding to the target crowd category.
In the embodiment of the present invention, the preset storage location may be a preset location for storing the target crowd category. Because the interesting contents of people in the same crowd category are similar, in the embodiment of the invention, the server stores the target crowd category to the preset storage position. The preset storage position may be located in a preset memory-based real-time storage system, where the memory-based real-time storage system may be a "touchbase" system, and data stored in the memory-based real-time storage system may be updated and acquired in real time. The historical logs in the preset duration can change, and the real-time logs can also change constantly, so that the stored target crowd categories often need to be updated, and when the terminal carries out video recommendation, the target crowd categories need to be read first. Therefore, in the embodiment of the invention, the preset storage position can be set in the real-time storage system based on the memory, and the target crowd category is stored in the real-time storage system, so that the target crowd category can be updated and acquired conveniently, and the video recommendation efficiency is improved. Specifically, during storage, the real-time storage system based on the memory may be stored in a preset storage location, that is, stored in the real-time storage system. The target crowd category may be composed of characters to facilitate storage and identification, the target time may be a time when the target crowd category is determined, and the memory-based real-time storage system may be deployed in the server or in another server, which is not limited in the embodiments of the present invention.
Accordingly, the terminal can quickly acquire the target crowd category through reading, and accordingly, the terminal can recommend the currently-viewed and identified user to be identified, namely the current user, based on the target recommendation strategy corresponding to the target crowd category. Because the user to be identified who is watching the video at present belongs to the target crowd category, when pushing is carried out according to the target crowd category, the recommended content can be ensured to be the content which is interested by the user to be identified who is watching the video at present, and then the recommending effect can be ensured. The target recommendation strategy can be a corresponding recommendation strategy formulated based on the content interested in the target crowd category, so as to ensure that the recommended content is the content interested by the current user.
Specifically, the terminal may read the target crowd category to which the current user belongs from the preset storage location under the condition that the current user interacts with the terminal through the target application. In the embodiment of the present invention, the interaction may be that the user clicks and starts the target application, or clicks and plays a video in the target application, or comments are made on the video in the target application, and the like. Correspondingly, if the user is detected to interact through the target application, the current user can be considered to start using the target application, and correspondingly, in order to improve the use experience of the current user, the terminal can read the target crowd category to which the current user belongs from the preset storage position of the server, so that video recommendation can be performed on the current user according to a recommendation strategy suitable for the current user. And then, recommending the content for the current user based on the target recommendation strategy corresponding to the target crowd category.
Further, since timeliness may affect the accuracy of the target demographic categories, for example, a user who uses a terminal in the morning may be different from a user who uses a terminal in the afternoon. Therefore, in the embodiment of the invention, the effectiveness of the target crowd category can be verified before the terminal reads the target crowd category. Specifically, the terminal may first determine whether a time interval between the target time and the current time is smaller than a preset interval threshold, and if the time interval between the target time and the current time is smaller than the preset interval threshold, the current target crowd category may be considered to be accurate, and therefore, the target crowd category may be directly read. On the contrary, if the time interval between the target time and the current time is not less than the preset interval threshold, the current target crowd category is considered to be not accurate enough. At this time, the update reminding information may be sent to the server so that the server updates the target crowd category, and accordingly, the terminal may read the updated target crowd category. Therefore, the server is controlled to update in time, the accuracy of the target crowd category can be ensured, and the effect of recommending based on the target crowd category is further ensured. The target time may be stored to a preset storage location when the server stores the target crowd category.
It should be noted that the viewing requirements and viewing habits of different people for videos are different, so that when the recommendation strategy is set, the display quantity and display mode of the recommended videos on each level of page can be defined in the recommendation strategy according to the viewing requirements and viewing habits of different people, and thus, when the videos are recommended to the user based on the target recommendation strategy, the use experience of the user can be further improved. For example, when recommendation is made for a group of children, no more than 6 videos may be exposed in a first-level category, no more than 6 videos may be exposed in the same last-level category, no more than 6 videos of the same type, that is, videos with the same tag, are provided, and the types of videos of the front brush and the rear brush which are continuously refreshed are the same. When the recommendation is carried out for non-child people, no more than 4 videos can be exposed in the first-level category, no more than 3 videos are exposed in the same last-level category, the videos of the same type, namely the videos with the same tag, are not more than 2, and the types of the videos which are continuously refreshed in a front brush and a back brush and are connected end to end are different.
Further, as time goes by, the crowd category to which the current user belongs may change, and accordingly, the terminal may switch the recommendation strategy corresponding to the target crowd category in the current situation according to different target crowd categories. For example, the terminal can switch from adopting a recommendation strategy for an adult group to a recommendation strategy for a child group, so as to ensure that the recommendation mode can meet the current user.
For example, fig. 2-2 is a schematic system diagram provided by an embodiment of the present invention, and taking the crowd categories including adults and children as an example, as shown in fig. 2-2, the crowd categories to which the historical users belong may be determined by an offline determination module, where the determination result may be "adults", "mixed" or "children", where "mixed" indicates that "adults" and "children" are included at the same time. Further, under the condition that the real-time behavior data comprise 'adults' and 'children', the target crowd category can be determined through the real-time recognition module according to the real-time behavior data. Further, as for the result of the real-time identification module, the real-time identification module may update the current time and the target time by using a "time limit" module shown in the figure, when the time length between the current time and the target time meets a preset time length threshold. The updating can also be performed by means of an "interaction" module shown in the figure, based on a new real-time log generated by the user's interaction with the target application. Of course, the results may also be stored directly by the "conventional" modules shown in the figures.
In summary, in the recommendation method provided in the embodiment of the present invention, when it is detected that a user to be identified watches a video, historical behavior data generated when the user to be identified watches the video in the past is obtained, a crowd category to which the user to be identified belongs is determined according to the historical behavior data, then, it is determined whether the crowd category to which the user to be identified belongs includes at least two crowd categories, if the crowd category to which the user to be identified belongs includes at least two crowd categories, current behavior data generated when the user to be identified watches the video at present is obtained, then, a target crowd category is determined according to the current behavior data, if the crowd category to which the user to be identified belongs does not include at least two crowd categories, the crowd category to which the user to be identified belongs is taken as the target crowd category, and finally, a target push content corresponding to the target crowd category is pushed to the user to be identified. Therefore, when at least two types of crowd categories are not included, the crowd category to which the user to be identified belongs is directly used as the target crowd category, so that the target crowd category is more accurate, and further the subsequent recommendation effect can be improved. Meanwhile, under the condition that different crowds are mixed and used in the process of watching the video in the past to cause damage to the portrait of the user, the recommended content can be ensured to be the content which is watched by the user to be identified and is interested by the user currently, and the recommending effect can be further ensured.
Fig. 3 is a block diagram of a recommendation apparatus according to an embodiment of the present invention, and as shown in fig. 3, the apparatus 30 may include:
the first obtaining module 301 is configured to, when it is detected that a user to be identified watches a video, obtain historical behavior data generated when the user to be identified watches the video in the past;
a first determining module 302, configured to determine, according to the historical behavior data, a crowd category to which the user to be identified belongs;
the judging module 303 is configured to judge whether the crowd categories to which the user to be identified belongs include at least two crowd categories;
a second obtaining module 304, configured to obtain current behavior data generated when the user to be identified watches a video currently if the crowd category to which the user to be identified belongs includes at least two crowd categories;
a second determining module 305, configured to determine a target crowd category according to the current behavior data;
a pushing module 306, configured to push, to the user to be identified, target push content corresponding to the target crowd category.
Optionally, the first obtaining module 301 is specifically configured to:
acquiring a history log generated by the application program within a preset time length from a history log file corresponding to a target application; the target application is an application program used by the user to be identified when watching a video;
for each historical log, extracting preset kinds of information from the historical log;
combining the preset types of information according to a preset information format to obtain the historical behavior data; the historical behavior data is used for representing specific information of various behaviors performed on the video by the historical user.
Optionally, the first determining module 302 is specifically configured to:
acquiring first preset judgment conditions corresponding to various preset crowd categories;
respectively matching the historical behavior data by utilizing each first preset judgment condition; the first preset judgment condition represents specific information required to be met by actions of a historical user on the video;
and determining a preset crowd category corresponding to a first preset judgment condition matched with the historical behavior data as the belonged crowd category.
Optionally, the second determining module 305 is specifically configured to:
determining a scene type to which the current scene belongs based on the current scene information and preset scene information corresponding to a plurality of preset scene types; the current scene information at least comprises the current time and/or place;
acquiring second preset judgment conditions corresponding to multiple preset crowd categories under the scene category;
matching the current behavior data by utilizing each second preset judgment condition; the second preset judgment condition represents specific information required to be met by the behavior executed by the current user on the video;
and determining a preset crowd category corresponding to a second preset judgment condition matched with the current behavior data as the target crowd category.
Optionally, the apparatus 30 further comprises:
the verification module is used for verifying the current behavior data based on preset verification conditions before the preset crowd category corresponding to the second preset judgment condition matched with the current behavior data is determined as the target crowd category;
and the execution module is used for executing the preset crowd category corresponding to the second preset judgment condition matched with the current behavior data if the verification is passed, and determining the preset crowd category as the operation of the target crowd category.
Optionally, the current behavior data is determined according to a real-time log generated by the target application; the device 30 further comprises:
the re-execution module is used for re-executing the step of determining the category of the target crowd through the current behavior data under the condition that the target application generates a new real-time log or the time length from the current moment to the target moment meets a preset time length threshold;
the target time is the time when the target crowd category is determined, and the real-time log is generated when the target application interacts with the user.
Optionally, the apparatus 30 further includes:
and the third determining module is used for taking the crowd category to which the user to be identified belongs as the target crowd category if the crowd category to which the user to be identified belongs does not comprise at least two crowd categories.
In summary, in the recommendation apparatus provided in the embodiment of the present invention, when it is detected that a user to be identified watches a video, a first obtaining module obtains historical behavior data generated when the user to be identified watches the video, a first determining module determines a crowd category to which the user to be identified belongs according to the historical behavior data, then, a determining module determines whether the crowd category to which the user to be identified belongs includes at least two crowd categories, if the crowd category to which the user to be identified belongs includes at least two crowd categories, a second obtaining module obtains current behavior data generated when the user to be identified watches the video currently, then, the second determining module determines a target crowd category according to the current behavior data, and finally, a pushing module pushes a target push content corresponding to the target crowd category to the user to be identified. Therefore, under the condition that the portrait of the user is damaged due to mixed use of different crowds in the process of watching the video in the past, the recommended content can be ensured to be the content which is interesting to the user to be identified and watching the video at present by adopting the recommending mode based on the target crowd category, and the recommending effect can be further ensured.
For the above device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and for the relevant points, refer to the partial description of the method embodiment.
An embodiment of the present invention further provides an electronic device, as shown in fig. 4, including a processor 401, a communication interface 402, a memory 403, and a communication bus 404, where the processor 401, the communication interface 402, and the memory 403 complete mutual communication through the communication bus 404,
a memory 403 for storing a computer program;
the processor 401, when executing the program stored in the memory 403, implements the following steps:
when a user to be identified watches a video, acquiring historical behavior data generated when the user to be identified watches the video in the past;
determining the crowd category to which the user to be identified belongs according to the historical behavior data;
judging whether the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
if the crowd categories to which the user to be identified belongs comprise at least two crowd categories, acquiring current behavior data generated when the user to be identified watches videos currently;
determining the target crowd category according to the current behavior data;
and pushing target push contents corresponding to the target crowd category to the user to be identified.
Wherein the target application is installed in the terminal.
The communication bus mentioned in the above terminal may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.
The communication interface is used for communication between the terminal and other equipment.
The Memory may include a Random Access Memory (RAM) or a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
In yet another embodiment of the present invention, a computer-readable storage medium is further provided, which has instructions stored therein, and when the instructions are executed on a computer, the instructions cause the computer to execute the recommendation method described in any of the above embodiments.
In yet another embodiment of the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform the recommendation method of any of the above embodiments.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), among others.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, as for the system embodiment, since it is substantially similar to the method embodiment, the description is relatively simple, and reference may be made to the partial description of the method embodiment for relevant points.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (14)
1. A recommendation method, comprising:
when it is detected that a user to be identified watches a video played by an application program, acquiring historical behavior data generated when the user to be identified watches the video in the past;
determining the crowd category to which the user to be identified belongs according to the historical behavior data;
judging whether the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
if the crowd categories to which the user to be identified belongs comprise at least two crowd categories, acquiring current behavior data generated when the user to be identified watches videos currently;
determining a target crowd type according to the current behavior data;
pushing target push contents corresponding to the target crowd category to the user to be identified;
determining a target crowd category through the current behavior data, wherein the determining the target crowd category comprises:
determining a scene type to which the current scene belongs based on the current scene information and preset scene information corresponding to a plurality of preset scene types; the current scene information at least comprises the current time and/or place;
acquiring second preset judgment conditions corresponding to multiple preset crowd categories under the scene category;
matching the current behavior data by using each second preset judgment condition; the second preset judgment condition represents specific information required to be met by the behavior executed by the current user on the video;
and determining a preset crowd category corresponding to a second preset judgment condition matched with the current behavior data as the target crowd category.
2. The method of claim 1, wherein the obtaining historical behavior data generated when the user to be identified watches videos in the past comprises:
acquiring a history log generated by the application program within a preset time length from a history log file corresponding to a target application; the target application is an application program used by the user to be identified when watching a video;
for each historical log, extracting preset kinds of information from the historical log;
combining the preset types of information according to a preset information format to obtain the historical behavior data; the historical behavior data is used for representing specific information of various behaviors performed on the video by historical users.
3. The method according to claim 1 or 2, wherein the determining the crowd category to which the user to be identified belongs according to the historical behavior data comprises:
acquiring first preset judgment conditions corresponding to various preset crowd categories;
matching the historical behavior data by utilizing each first preset judgment condition; the first preset judgment condition represents specific information required to be met by actions of historical users on the video;
and determining a preset crowd category corresponding to a first preset judgment condition matched with the historical behavior data as the belonged crowd category.
4. The method of claim 1, further comprising:
verifying the current behavior data based on preset verification conditions before determining a preset crowd category corresponding to a second preset judgment condition matched with the current behavior data as the target crowd category;
and if the current behavior data passes the verification, executing the preset crowd category corresponding to the second preset judgment condition matched with the current behavior data, and determining the preset crowd category as the target crowd category.
5. The method of claim 2, wherein the current behavior data is determined from a real-time log generated by the target application; the method further comprises the following steps:
a step of determining the category of the target crowd through the current behavior data when the target application generates a new real-time log or when the time length from the current moment to the target moment meets a preset time length threshold;
the target time is the time when the target crowd category is determined, and the real-time log is generated when the target application interacts with the user.
6. The method of claim 1, wherein after determining whether the crowd categories to which the user to be identified belongs include at least two crowd categories, the method further comprises:
and if the crowd category to which the user to be identified belongs does not comprise at least two crowd categories, taking the crowd category to which the user to be identified belongs as the target crowd category.
7. A recommendation device, comprising:
the device comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring historical behavior data generated when a user to be identified watches a video played by an application program and the user to be identified watches the video;
the first determining module is used for determining the crowd category to which the user to be identified belongs according to the historical behavior data;
the judging module is used for judging whether the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
the second obtaining module is used for obtaining current behavior data generated when the user to be identified watches the video currently if the crowd categories to which the user to be identified belongs comprise at least two crowd categories;
the second determining module is used for determining the target crowd category according to the current behavior data;
the pushing module is used for pushing target pushing content corresponding to the target crowd category to the user to be identified;
the second determining module is specifically configured to:
determining a scene type to which the current scene belongs based on the current scene information and preset scene information corresponding to a plurality of preset scene types; the current scene information at least comprises the current time and/or place;
acquiring second preset judgment conditions corresponding to multiple preset crowd categories under the scene category;
matching the current behavior data by using each second preset judgment condition; the second preset judgment condition represents specific information required to be met by the behavior executed by the current user on the video;
and determining a preset crowd category corresponding to a second preset judgment condition matched with the current behavior data as the target crowd category.
8. The apparatus of claim 7, wherein the first obtaining module is specifically configured to:
acquiring a historical log generated by the application program within a preset time length from a historical log file corresponding to a target application; the target application is an application program used by the user to be identified when watching the video;
for each historical log, extracting preset types of information from the historical log;
combining the preset types of information according to a preset information format to obtain the historical behavior data; the historical behavior data is used for representing specific information of various behaviors performed on the video by historical users.
9. The apparatus according to claim 7 or 8, wherein the first determining module is specifically configured to:
acquiring first preset judgment conditions corresponding to various preset crowd categories;
respectively matching the historical behavior data by utilizing each first preset judgment condition; the first preset judgment condition represents specific information required to be met by actions of a historical user on the video;
and determining a preset crowd category corresponding to a first preset judgment condition matched with the historical behavior data as the belonged crowd category.
10. The apparatus of claim 7, further comprising:
the verification module is used for verifying the current behavior data based on preset verification conditions before determining a preset crowd category corresponding to a second preset judgment condition matched with the current behavior data as the target crowd category;
and the execution module is used for executing the preset crowd category corresponding to the second preset judgment condition matched with the current behavior data if the verification is passed, and determining the preset crowd category as the target crowd category.
11. The apparatus of claim 8, wherein the current behavior data is determined from a real-time log generated by the target application; the device further comprises:
the re-execution module is used for re-executing the step of determining the category of the target crowd through the current behavior data under the condition that the target application generates a new real-time log or the time length from the current moment to the target moment meets a preset time length threshold;
the target time is the time when the target crowd category is determined, and the real-time log is generated when the target application interacts with the user.
12. The apparatus of claim 7, further comprising:
and the third determining module is used for taking the crowd category to which the user to be identified belongs as the target crowd category if the crowd category to which the user to be identified belongs does not comprise at least two crowd categories.
13. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any one of claims 1 to 6 when executing a program stored in a memory.
14. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911184565.9A CN110941738B (en) | 2019-11-27 | 2019-11-27 | Recommendation method and device, electronic equipment and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911184565.9A CN110941738B (en) | 2019-11-27 | 2019-11-27 | Recommendation method and device, electronic equipment and computer-readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110941738A CN110941738A (en) | 2020-03-31 |
| CN110941738B true CN110941738B (en) | 2023-03-28 |
Family
ID=69908375
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911184565.9A Active CN110941738B (en) | 2019-11-27 | 2019-11-27 | Recommendation method and device, electronic equipment and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110941738B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112115298B (en) * | 2020-09-15 | 2023-07-25 | 北京奇艺世纪科技有限公司 | Video recommendation method and device, electronic equipment and storage medium |
| CN112734486A (en) * | 2021-01-13 | 2021-04-30 | 北京明略昭辉科技有限公司 | Method and device for determining target crowd, electronic equipment and storage medium |
| CN114374881B (en) * | 2022-01-05 | 2023-09-01 | 北京百度网讯科技有限公司 | Method and device for distributing user traffic, electronic equipment and storage medium |
| CN114201741B (en) * | 2022-02-18 | 2022-05-06 | 北京派瑞威行互联技术有限公司 | Method, device and machine-readable storage medium for information processing |
| CN114926771A (en) * | 2022-06-01 | 2022-08-19 | 北京字节跳动网络技术有限公司 | Video identification method and device |
| CN115037957B (en) * | 2022-06-07 | 2024-01-30 | 北京视达科技有限公司 | Method, device and system for recommending on-demand content based on live program |
| CN116567306B (en) * | 2023-05-09 | 2023-10-20 | 北京新东方迅程网络科技有限公司 | Video recommendation method and device, electronic equipment and medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105681835A (en) * | 2016-02-26 | 2016-06-15 | 腾讯科技（深圳）有限公司 | Information pushing method and server |
| CN105898413A (en) * | 2016-05-24 | 2016-08-24 | 青岛海信电器股份有限公司 | Television program recommendation method, television and recommendation server |
| CN106162348A (en) * | 2015-04-13 | 2016-11-23 | 海信集团有限公司 | A kind of personal program recommends method and device |
| CN108959634A (en) * | 2018-07-25 | 2018-12-07 | 南京创维信息技术研究院有限公司 | Video recommendation method, device, equipment and storage medium |
| CN109033149A (en) * | 2018-06-12 | 2018-12-18 | 北京奇艺世纪科技有限公司 | Information recommendation method, device, server and storage medium |
| CN109871756A (en) * | 2019-01-10 | 2019-06-11 | 平安科技（深圳）有限公司 | The method, apparatus and computer equipment of user tag are established based on recognition of face |
| CN110035302A (en) * | 2018-01-11 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Information recommendation and model training method and device calculate equipment, storage medium |
| CN110087103A (en) * | 2019-04-26 | 2019-08-02 | 北京奇艺世纪科技有限公司 | A kind of video recommendation system, method, apparatus and computer |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10601953B2 (en) * | 2016-09-22 | 2020-03-24 | Adobe Inc. | Decomposing media content accounts for persona-based experience individualization |
-
2019
- 2019-11-27 CN CN201911184565.9A patent/CN110941738B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106162348A (en) * | 2015-04-13 | 2016-11-23 | 海信集团有限公司 | A kind of personal program recommends method and device |
| CN105681835A (en) * | 2016-02-26 | 2016-06-15 | 腾讯科技（深圳）有限公司 | Information pushing method and server |
| CN105898413A (en) * | 2016-05-24 | 2016-08-24 | 青岛海信电器股份有限公司 | Television program recommendation method, television and recommendation server |
| CN110035302A (en) * | 2018-01-11 | 2019-07-19 | 阿里巴巴集团控股有限公司 | Information recommendation and model training method and device calculate equipment, storage medium |
| CN109033149A (en) * | 2018-06-12 | 2018-12-18 | 北京奇艺世纪科技有限公司 | Information recommendation method, device, server and storage medium |
| CN108959634A (en) * | 2018-07-25 | 2018-12-07 | 南京创维信息技术研究院有限公司 | Video recommendation method, device, equipment and storage medium |
| CN109871756A (en) * | 2019-01-10 | 2019-06-11 | 平安科技（深圳）有限公司 | The method, apparatus and computer equipment of user tag are established based on recognition of face |
| CN110087103A (en) * | 2019-04-26 | 2019-08-02 | 北京奇艺世纪科技有限公司 | A kind of video recommendation system, method, apparatus and computer |
Non-Patent Citations (1)
| Title |
|---|
| "一种面向共享账号的个性化推荐算法";李伟 等;《计算机应用研究》;20171010;第35卷(第10期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110941738A (en) | 2020-03-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110941738B (en) | Recommendation method and device, electronic equipment and computer-readable storage medium | |
| KR101816113B1 (en) | Estimating and displaying social interest in time-based media | |
| US9892109B2 (en) | Automatically coding fact check results in a web page | |
| CN106331778B (en) | Video recommendation method and device | |
| CN106649316B (en) | Video pushing method and device | |
| WO2018036272A1 (en) | News content pushing method, electronic device, and computer readable storage medium | |
| CN110475155B (en) | Live video hot state identification method, device, equipment and readable medium | |
| US10348550B2 (en) | Method and system for processing network media information | |
| WO2017101506A1 (en) | Information processing method and device | |
| CN109635155B (en) | Method and device for pushing video to user, electronic equipment and storage medium | |
| CN109511015B (en) | Multimedia resource recommendation method, device, storage medium and equipment | |
| CN112364202B (en) | Video recommendation method and device and electronic equipment | |
| CN110035302B (en) | Information recommendation method and device, model training method and device, computing equipment and storage medium | |
| US20170169062A1 (en) | Method and electronic device for recommending video | |
| CN111479129A (en) | Live broadcast cover determining method, device, server, medium and system | |
| CN112199582A (en) | Content recommendation method, device, equipment and medium | |
| CN112995690A (en) | Live content item identification method and device, electronic equipment and readable storage medium | |
| CN110121088B (en) | User attribute information determination method and device and electronic equipment | |
| CN110427358B (en) | Data cleaning method and device and information recommendation method and device | |
| CN111314790A (en) | Video playing record sequencing method and device and electronic equipment | |
| CN111314744A (en) | Video pushing method and server | |
| CN111556326A (en) | Public class video clip pushing method and device, electronic equipment and storage medium | |
| CN110971973A (en) | Video pushing method and device and electronic equipment | |
| CN110569447A (en) | network resource recommendation method and device and storage medium | |
| CN112507214B (en) | User name-based data processing method, device, equipment and medium |
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 |