Association application and data analysis method based on audio and video data

Abstract

The invention discloses an associated application and data analysis method based on audio and video data, and the analysis method comprises the steps: obtaining an application feature of a target application in response to an analysis demand; the application features comprise a plurality of application functions; based on the application features of the target application, obtaining at least one associated application and target audio and video data of each associated application; obtaining a target topic and topic feature data based on the target audio and video data of the plurality of associated applications; the topic feature data comprises search data, browsing data and comment data; and obtaining a recommendation strategy of the target application based on the target topics and the topic feature data of the plurality of associated applications. The technical problem that in the prior art, personalized recommendation service efficiency and accuracy of applications are low is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of audio and video data analysis, and in particular to an associated application and data analysis method based on audio and video data. Background Art

[0002] With the development of Internet technology, users browse a large amount of audio and video data in their daily lives. Therefore, accurate analysis of the topics of audio and video data is crucial for information screening, reading, content retrieval, and personalized recommendations.

[0003] In the prior art, applications collect and analyze the audio and video data that users browse in the application, obtain topics that users are interested in, and make personalized recommendations. However, the data between each application is independent of each other. Users need to set the content they are interested in in the new application, or a certain amount of audio and video data must be generated in the new application before the new application can filter and make personalized recommendations.

[0004] It can be seen that in the process of using the application, how to quickly and accurately provide personalized recommendation services is an urgent problem to be solved. Summary of the invention

[0005] The purpose of the present invention is to provide an associated application and a data analysis method based on audio and video data, which solves the technical problem of low efficiency and accuracy of personalized recommendation services applied in the prior art.

[0006] The present invention provides an associated application based on audio and video data and a data analysis method, the analysis method comprising:

[0007] Respond to analysis requirements and obtain application characteristics of the target application; application characteristics include several application functions;

[0008] Based on the application characteristics of the target application, obtaining at least one associated application and target audio and video data of each associated application;

[0009] Based on the target audio and video data of several related applications, the target topic and topic feature data are obtained; the topic feature data includes search data, browsing data and comment data;

[0010] Based on target topics and topic feature data of several related applications, a recommendation strategy for the target application is obtained.

[0011] Further, based on the application characteristics of the target application, obtaining at least one associated application and target audio and video data of each associated application includes:

[0012] Obtain multiple historical applications installed in the current user equipment terminal, and application features corresponding to each historical application;

[0013] Based on the application characteristics of the target application, a number of related applications and a related function corresponding to each related application are obtained from a number of historical applications;

[0014] Based on the associated function of each associated application, target audio and video data corresponding to the associated function in each associated application is obtained.

[0015] Furthermore, based on the application characteristics of the target application, a number of related applications and a related function corresponding to each related application are obtained from a number of historical applications, including:

[0016] Based on several application functions of the target application, obtaining an associated directory; the associated directory includes several applications and corresponding application functions; each application function of the target application is associated with an application function of at least one application;

[0017] When any application function of a historical application is in the associated directory, the historical application is an associated application of the target application, and the application function is an associated function.

[0018] Furthermore, obtaining target audio and video data of the associated application includes:

[0019] Obtaining several associated functions of associated applications;

[0020] Based on each associated function, audio and video data in the associated application is obtained as target audio and video data.

[0021] Furthermore, based on the target audio and video data of several related applications, target topics and topic feature data are obtained, including:

[0022] Obtain a number of target audio and video data corresponding to each application function in the target application;

[0023] Based on a number of target audio and video data corresponding to an application function of a target application, a target topic corresponding to the application function and search data, browsing data and comment data of users under the target topic are obtained as topic feature data;

[0024] The search data includes the number of searches and search frequency; the browsing data includes the number of browses, browsing time and browsing frequency; the comment data includes the number of comments, average comment word count and comment duration.

[0025] Further, based on a number of target audio and video data corresponding to an application function of the target application, a target topic corresponding to the application function is obtained, including:

[0026] Based on a number of target audio and video data, obtaining target text data;

[0027] Based on the target text data, obtain keyword data;

[0028] Based on keyword data, confirm the target topic.

[0029] Furthermore, based on the target topics and topic feature data of several related applications, a recommendation strategy for the target application is obtained, including:

[0030] Obtain target topics and topic feature data corresponding to each application function of the target application;

[0031] Based on the target topic and topic feature data corresponding to each application function, obtaining a recommendation sub-strategy corresponding to each application function; the recommendation sub-strategy includes topic recommendation time and topic recommendation priority;

[0032] Based on the recommendation sub-strategies corresponding to several application functions, a recommendation strategy for a target application is obtained.

[0033] Furthermore, based on the target topic and topic feature data corresponding to each application function, a recommendation sub-strategy corresponding to each application function is obtained, including:

[0034] Acquire a strategic relationship database, wherein the strategic relationship database includes target topics, search data, browsing data, comment data, topic recommendation time, topic recommendation priority, and mapping relationships between target topics, search data, browsing data, comment data, topic recommendation time, and topic recommendation priority;

[0035] Train the neural network model through the strategic relationship database to obtain the strategic model;

[0036] The target topic and topic feature data corresponding to the application function are input into the strategy model to obtain the recommended sub-strategy.

[0037] Furthermore, the recommendation sub-strategy also includes:

[0038] Obtain several target topics corresponding to the application functions, as well as the topic recommendation time and topic recommendation priority of each target topic;

[0039] Based on the topic recommendation time of several target topics, several recommended time periods are obtained;

[0040] Based on the topic recommendation priority of each target topic in the same recommendation period, the recommendation proportion of each target topic in the preset display range is obtained.

[0041] Furthermore, the preset display range includes:

[0042] The maximum range displayed on the display screen of the device terminal when the user performs several preset gestures on the device terminal continuously.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] In the present invention, by obtaining at least one associated application of the target application according to the application characteristics of the target application, the associated applications of the target application are obtained more accurately; by obtaining the target topic and topic feature data according to the target audio and video data corresponding to each associated application, the analysis of topics of interest to the user is made more accurate. By analyzing the target topic and topic feature data, the recommendation strategy is obtained, which effectively improves the efficiency and accuracy of content screening and personalized recommendation of the target application. The technical problem of low efficiency and accuracy of personalized recommendation services applied in the prior art is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The present invention is a principle block diagram of the associated application and data analysis method based on audio and video data. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0047] like Figure 1 As shown, based on the associated application of audio and video data and the data analysis method, the analysis method includes:

[0048] Step S10: In response to the analysis requirement, the application characteristics of the target application are obtained; the application characteristics include several application functions;

[0049] In this embodiment, after the target application is installed, the device terminal pops up an analysis requirement for the user to select; after receiving the user's consent instruction, the device terminal responds to the analysis requirement. The application characteristics of the target application are obtained.

[0050] In this embodiment, the application features include several application functions; the several application functions include short videos, medium videos, TV series, music, shopping, etc. In this embodiment, each target application includes at least one application function.

[0051] Step S20: based on the application characteristics of the target application, obtaining at least one associated application and target audio and video data of each associated application;

[0052] In this embodiment, according to the application characteristics of the target application, several historical applications in the device terminal are obtained, and according to the application characteristics of the target application, one or more associated applications are obtained from the several historical applications. In this embodiment, the device terminal includes a smart phone, a tablet computer, a laptop computer, and a desktop computer. The historical applications include applications currently installed in the device terminal and / or applications that have been installed in the device terminal in the past.

[0053] Step S30: based on the target audio and video data of several related applications, obtaining target topics and topic feature data; the topic feature data includes search data, browsing data and comment data;

[0054] In this embodiment, the target audio and video data is analyzed to obtain the target topic and topic feature data, thereby reducing the process of users generating audio and video data in the target application before personalized recommendations can be made.

[0055] Step S40: Based on the target topics and topic feature data of several related applications, a recommendation strategy for the target application is obtained.

[0056] In this embodiment, the target topic and topic feature data of several related applications are analyzed to obtain the recommendation strategy corresponding to the target application, thereby improving the accuracy of personalized recommendation of the target application, enabling the target application to quickly adapt to the user and improving the user experience.

[0057] The specific implementation process of this embodiment includes:

[0058] In this embodiment, by obtaining at least one associated application of the target application according to the application characteristics of the target application, the associated applications of the target application are obtained more accurately; by obtaining the target topic and topic feature data according to the target audio and video data corresponding to each associated application, the analysis of topics of interest to the user is made more accurate. By analyzing the target topic and topic feature data and obtaining the recommendation strategy, the efficiency and accuracy of content screening and personalized recommendation of the target application are effectively improved. The technical problem of low efficiency and accuracy of personalized recommendation services applied in the prior art is solved.

[0059] In this embodiment, based on the application characteristics of the target application, obtaining at least one associated application and target audio and video data of each associated application includes:

[0060] Obtain multiple historical applications installed in the current user equipment terminal, and application features corresponding to each historical application;

[0061] Based on the application characteristics of the target application, a number of related applications and a related function corresponding to each related application are obtained from a number of historical applications;

[0062] Based on the associated function of each associated application, target audio and video data corresponding to the associated function in each associated application is obtained.

[0063] In this embodiment, the application characteristics of the historical application include several application functions, and the several application functions include short videos, medium videos, TV series, music, shopping, etc. It should be noted that the application characteristics of the historical application in this embodiment may be consistent with the target application, or inconsistent with the target application, and it is necessary to determine whether the historical application is associated with the target application. If so, the historical application is an associated application of the target application; if not, the historical application is a non-associated application of the target application. The audio and video data in the non-associated application has no reference significance for the personalized recommendation content of the target application, or the reference significance is small, and the audio and video data of the non-associated application will cause the personalized recommendation of the target application to be inaccurate. Therefore, in this embodiment, by obtaining and analyzing the audio and video data in the associated application, the effect of personalized recommendation of the target application can be improved.

[0064] In this embodiment, for associated applications, not all audio and video data corresponding to functions are of reference significance. Only audio and video data corresponding to associated functions in associated applications are of reference significance for personalized recommendation content of application functions of the target application.

[0065] The audio and video data corresponding to the non-related functions in the related applications have no reference significance for the personalized recommendation content of the application functions of the target application. If the audio and video data corresponding to the non-related functions are analyzed, the personalized recommendation content of the application functions of the target application will be inaccurate, affecting the user experience.

[0066] In this embodiment, based on the application characteristics of the target application, a number of associated applications and associated functions corresponding to each associated application are obtained from a number of historical applications, including:

[0067] Based on several application functions of the target application, obtaining an associated directory; the associated directory includes several applications and corresponding application functions; each application function of the target application is associated with an application function of at least one application;

[0068] When any application function of a historical application is in the associated directory, the historical application is an associated application of the target application, and the application function is an associated function.

[0069] In this embodiment, when developing various application functions in the target application, the target application development experts analyze various applications on the market and the application functions therein, and generate a related directory based on their experience. According to the related directory, it can be determined whether the historical application is a related application, and the related functions of the related application can be obtained.

[0070] In this embodiment, obtaining target audio and video data of the associated application includes:

[0071] Obtaining several associated functions of associated applications;

[0072] Based on each associated function, audio and video data in the associated application is obtained as target audio and video data.

[0073] The audio and video data corresponding to all the associated functions in the associated application are obtained as the target audio and video data, and the audio and video data corresponding to the non-associated functions are ignored.

[0074] In this embodiment, based on the target audio and video data of several related applications, the target topic and topic feature data are obtained, including:

[0075] Obtain a number of target audio and video data corresponding to each application function in the target application;

[0076] Based on a number of target audio and video data corresponding to an application function of a target application, a target topic corresponding to the application function and search data, browsing data and comment data of users under the target topic are obtained as topic feature data;

[0077] The search data includes the number of searches and search frequency; the browsing data includes the number of browses, browsing time and browsing frequency; the comment data includes the number of comments, average comment word count and comment duration.

[0078] In this embodiment, by acquiring a number of target audio and video data corresponding to each application function, the target topic and topic feature data corresponding to each application function are analyzed and acquired. In this embodiment, the topic feature data includes search data, browsing data and comment data as topic feature data.

[0079] It should be noted that, in this embodiment, each application function corresponds to at least one target topic, and search data, browsing data, and comment data under each target topic are obtained as topic feature data.

[0080] In this embodiment, based on a number of target audio and video data corresponding to an application function of a target application, obtaining a target topic corresponding to the application function includes:

[0081] Based on a number of target audio and video data, obtaining target text data;

[0082] Based on the target text data, obtain keyword data;

[0083] Based on keyword data, confirm the target topic.

[0084] The specific implementation process of this embodiment includes:

[0085] The target audio and video files are processed through the text extraction model to obtain the target text data; the text data is processed through the keyword recognition model to obtain the keywords in the target text data to generate keyword data; the keyword data is processed through the topic analysis model to obtain the target topic.

[0086] In this embodiment, the text extraction model, keyword recognition model and topic analysis model are all convolutional neural network models.

[0087] In this embodiment, based on the target topics and topic feature data of several related applications, a recommendation strategy for the target application is obtained, including:

[0088] Obtain target topics and topic feature data corresponding to each application function of the target application;

[0089] Based on the target topic and topic feature data corresponding to each application function, obtaining a recommendation sub-strategy corresponding to each application function; the recommendation sub-strategy includes topic recommendation time and topic recommendation priority;

[0090] Based on the recommendation sub-strategies corresponding to several application functions, a recommendation strategy for a target application is obtained.

[0091] In this embodiment, a recommendation sub-strategy corresponding to each application function is generated according to the target topic and topic feature data corresponding to each application function; this can make the personalized recommendation content of each application function in the target application more accurate and improve the user experience.

[0092] In this embodiment, the recommendation sub-strategies corresponding to several application functions constitute the recommendation strategy of the target application.

[0093] In this embodiment, based on the target topic and topic feature data corresponding to each application function, the recommendation sub-strategy corresponding to each application function is obtained, including:

[0094] Acquire a strategic relationship database, wherein the strategic relationship database includes target topics, search data, browsing data, comment data, topic recommendation time, topic recommendation priority, and mapping relationships between target topics, search data, browsing data, comment data, topic recommendation time, and topic recommendation priority;

[0095] Train the neural network model through the strategic relationship database to obtain the strategic model;

[0096] The target topic and topic feature data corresponding to the application function are input into the strategy model to obtain the recommended sub-strategy.

[0097] In this embodiment, during the training of the neural network model, a training set and a validation set are obtained according to the policy relationship database, the neural network model is trained by the validation set, and the policy model is obtained after being verified by the validation set.

[0098] According to another embodiment of the present invention, the recommendation sub-strategy further includes:

[0099] Obtain several target topics corresponding to the application functions, as well as the topic recommendation time and topic recommendation priority of each target topic;

[0100] Based on the topic recommendation time of several target topics, several recommended time periods are obtained;

[0101] Based on the topic recommendation priority of each target topic in the same recommendation period, the recommendation proportion of each target topic in the preset display range is obtained.

[0102] The specific implementation process of this embodiment includes:

[0103] In this embodiment, the priority of topic recommendation is A, B, C, and D from high to low;

[0104] Each topic recommendation priority is set with a corresponding recommendation ratio. In this embodiment, when the topic recommendation priority is level A, the recommendation ratio is 4; when the topic recommendation priority is level B, the recommendation ratio is 3; when the topic recommendation priority is level C, the recommendation ratio is 2; when the topic recommendation priority is level D, the recommendation ratio is 1.

[0105] According to the recommendation ratio corresponding to each target topic, obtain the recommendation ratio of each target topic. The calculation formula is as follows:

[0106]

[0107] Among them, ZB i is the recommendation ratio of the i-th target topic; TB i is the recommendation ratio of the i-th target topic; n is the total number of target topics.

[0108] In this embodiment, the preset display range includes:

[0109] The maximum range displayed on the display screen of the device terminal when the user performs several preset gestures on the device terminal continuously.

[0110] In this embodiment, the preset gestures include sliding the screen, rolling the mouse, and clicking a virtual button on the real screen.

[0111] For example, for a smartphone or tablet computer, the preset display range includes the maximum range displayed on the display screen when the screen is swiped three times in succession;

[0112] For desktop computers, the preset display range includes the maximum range displayed on the screen when the mouse is scrolled three times in succession.

[0113] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0114] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An application and data analysis method based on audio and video data, characterized in that: The analysis method comprises: Respond to analysis requirements and obtain application characteristics of the target application; application characteristics include several application functions; Based on the application characteristics of the target application, obtaining at least one associated application and target audio and video data of each associated application; Based on the target audio and video data of several related applications, the target topic and topic feature data are obtained; the topic feature data includes search data, browsing data and comment data; Based on target topics and topic feature data of several related applications, a recommendation strategy for the target application is obtained.

2. The method for associating audio and video data and analyzing data according to claim 1, characterized in that: Based on the application characteristics of the target application, obtaining at least one associated application and target audio and video data of each associated application includes: Obtain multiple historical applications installed in the current user equipment terminal, and application features corresponding to each historical application; Based on the application characteristics of the target application, a number of related applications and a related function corresponding to each related application are obtained from a number of historical applications; Based on the associated function of each associated application, target audio and video data corresponding to the associated function in each associated application is obtained.

3. The method for associating audio and video data and analyzing data according to claim 2, characterized in that: Based on the application characteristics of the target application, a number of related applications and related functions corresponding to each related application are obtained from a number of historical applications, including: Based on several application functions of the target application, obtaining an associated directory; the associated directory includes several applications and corresponding application functions; each application function of the target application is associated with an application function of at least one application; When any application function of a historical application is in the associated directory, the historical application is an associated application of the target application, and the application function is an associated function.

4. The method for associating audio and video data and analyzing data according to claim 3, characterized in that: Obtaining target audio and video data of associated applications includes: Obtaining several associated functions of associated applications; Based on each associated function, audio and video data in the associated application is obtained as target audio and video data.

5. The method for associating audio and video data and analyzing data according to claim 4, characterized in that: Based on the target audio and video data of several related applications, the target topic and topic feature data are obtained, including: Obtain a number of target audio and video data corresponding to each application function in the target application; Based on a number of target audio and video data corresponding to an application function of a target application, a target topic corresponding to the application function and search data, browsing data and comment data of users under the target topic are obtained as topic feature data; The search data includes the number of searches and search frequency; the browsing data includes the number of browses, browsing time and browsing frequency; the comment data includes the number of comments, average comment word count and comment duration.

6. The method for associating audio and video data and analyzing data according to claim 5, characterized in that: Based on a plurality of target audio and video data corresponding to an application function of a target application, obtaining a target topic corresponding to the application function includes: Based on a number of target audio and video data, obtaining target text data; Based on the target text data, obtain keyword data; Based on keyword data, confirm the target topic.

7. The method for associating audio and video data and analyzing data according to claim 5, characterized in that: Based on the target topics and topic feature data of several related applications, a recommendation strategy for the target application is obtained, including: Obtain target topics and topic feature data corresponding to each application function of the target application; Based on the target topic and topic feature data corresponding to each application function, obtaining a recommendation sub-strategy corresponding to each application function; the recommendation sub-strategy includes topic recommendation time and topic recommendation priority; Based on the recommendation sub-strategies corresponding to several application functions, a recommendation strategy for a target application is obtained.

8. The method for associating audio and video data and analyzing data according to claim 7, characterized in that: Based on the target topic and topic feature data corresponding to each application function, obtain the recommendation sub-strategy corresponding to each application function, including: Acquire a strategic relationship database, wherein the strategic relationship database includes target topics, search data, browsing data, comment data, topic recommendation time, topic recommendation priority, and mapping relationships between target topics, search data, browsing data, comment data, topic recommendation time, and topic recommendation priority; Train the neural network model through the strategic relationship database to obtain the strategic model; The target topic and topic feature data corresponding to the application function are input into the strategy model to obtain the recommended sub-strategy.

9. The method for associating audio and video data and analyzing data according to claim 8, characterized in that: The recommendation sub-strategy also includes: Obtain several target topics corresponding to the application functions, as well as the topic recommendation time and topic recommendation priority of each target topic; Based on the topic recommendation time of several target topics, several recommended time periods are obtained; Based on the topic recommendation priority of each target topic in the same recommendation period, the recommendation proportion of each target topic in the preset display range is obtained.

10. The method for associating audio and video data and analyzing data according to claim 9, characterized in that: The preset display range includes: The maximum range displayed on the display screen of the device terminal when the user performs several preset gestures on the device terminal continuously.

Images