List generation method and device

Abstract

The embodiment of the invention provides a list generation method, and the method comprises the steps: obtaining live broadcast interaction data of each live broadcast room, and determining the content type of each live broadcast room; a scoring factor corresponding to each content type is obtained, different content types correspond to different scoring factors, and the scoring factors comprise the weight of the live broadcast interaction data; determining the score of each live broadcast room based on the live broadcast interaction data and the score factor; and generating a list based on the score of each live broadcast room. According to the technical scheme provided by the embodiment of the invention, the generated list can reasonably reflect the popularity degree of the live broadcast content, and the list generation accuracy is improved.

Description

Technical Field

[0001] Embodiments of the present application relate to the technical field of data processing, and in particular, to a method, device, computer device, computer-readable storage medium, and computer program product for generating a list. Background Art

[0002] Currently, when many live streaming platforms generate a popularity list, they usually rely on data such as the number of online viewers, barrage data, and viewing duration corresponding to the live content to generate the list. This method of generating a list often results in live content with a high viewing duration (such as movies, TV dramas, etc.) occupying the top positions in the list. However, this live content with a high viewing duration generally has low interactivity and is not necessarily the most popular live content. It only ranks at the top of the list because of the long viewing duration; while other live content with high interactivity is usually the most popular live content, but it cannot rank at the top of the list because of the short live duration. Therefore, the current method of generating a list often cannot accurately reflect the popularity of the live content, and the accuracy is low.

[0003] It should be noted that the above content is not necessarily prior art and is not used to limit the patent protection scope of the present application. Summary of the Invention

[0004] Embodiments of the present application provide a method, device, computer device, computer-readable storage medium, and computer program product for generating a list to solve or alleviate one or more of the above technical problems.

[0005] One aspect of the embodiments of the present application provides a method for generating a list, the method including: Obtaining the live interaction data of each live room and determining the content type of each live room; Obtaining a scoring factor corresponding to each content type, where different content types correspond to different scoring factors, and the scoring factor includes the weight of the live interaction data; Determining the score of each live room based on the live interaction data and the scoring factor; Generating a list based on the score of each live room.

[0006] Optionally, the determining the content type of each live room includes: Obtaining the target data of each live room, where the target data includes at least part of the data of the title, description, tags, and picture content of the live room; Determining the content type of each live room according to the target data by using a deep learning model and / or natural language processing method.

[0007] Optionally, determining the score of each live room based on the live interaction data and the scoring factor includes: Obtain the target interaction data in the live interaction data; Determine the interaction quality of each live room based on the target interaction data; Determine the score of each live room based on the live interaction data, the scoring factor, and the interaction quality.

[0008] Optionally, determining the interaction quality of each live room based on the target interaction data includes: Perform sentiment analysis and semantic depth analysis on the target interaction data using natural language processing methods to obtain a sentiment analysis result and a semantic depth analysis result; Determine the interaction quality of each live room based on the sentiment analysis result and the semantic depth analysis result.

[0009] Optionally, the content type includes a target content type, the duration of the target content type is greater than a preset threshold, and the scoring factor corresponding to the target content type further includes a duration penalty factor.

[0010] Optionally, the method further includes: Obtain the historical interaction data of the current user when the current user authorizes; Correspondingly, determining the score of each live room based on the live interaction data and the scoring factor includes: Adjust the scoring factor based on the historical interaction data to obtain an adjusted scoring factor; Determine the score of each live room based on the live interaction data and the adjusted scoring factor.

[0011] Another aspect of the embodiments of the present application provides a list generation device, and the device includes: A determination module, configured to obtain the live interaction data of each live room and determine the content type of each live room; An acquisition module, configured to acquire a scoring factor corresponding to each content type, where different content types correspond to different scoring factors, and the scoring factor includes the weight of the live interaction data; A scoring module, configured to determine the score of each live room based on the live interaction data and the scoring factor; A generation module, configured to generate a list based on the score of each live room.

[0012] Another aspect of the embodiments of the present application provides a computer device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein: the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the method as described above.

[0013] Another aspect of the embodiments of the present application provides a computer-readable storage medium, in which computer instructions are stored, and when the computer instructions are executed by a processor, the method as described above is implemented.

[0014] Another aspect of the embodiments of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, the method as described above is implemented.

[0015] The embodiments of the present application adopting the above technical solutions may include the following advantages: By obtaining the live interaction data of each live broadcast room, determining the content type of each live broadcast room, obtaining the scoring factors corresponding to each content type, determining the score of each live broadcast room based on the live interaction data and the scoring factors, and generating a list based on the scores of each live broadcast room. Since different content types correspond to different scoring factors, and the scoring factors include the weights of the live interaction data, it is possible to score the live broadcast rooms of different content types according to different scoring rules, so that the generated list can reasonably reflect the popularity of the live content and improve the accuracy of list generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings exemplarily show the embodiments and form a part of the specification, and are used together with the written description of the specification to explain the exemplary embodiments. The shown embodiments are only for the purpose of illustration and do not limit the scope of the claims. In all the drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0017] Figure 1 Schematically shows the operating environment diagram of the list generation method according to Embodiment 1 of the present application; Figure 2 Schematically shows the flowchart of the list generation method according to Embodiment 1 of the present application; Figure 3 Schematically shows Figure 2 the sub-step flowchart of step S200 in; Figure 4 Schematically shows Figure 2 the sub-step flowchart of step S204 in; Figure 5 Schematically shows Figure 4The flowchart of the sub-steps of step S402; Figure 6 Schematically shows Figure 2 Another flowchart of the sub-steps of step S204; Figure 7 Schematically shows the block diagram of the list generation device according to Embodiment 2 of the present application; and Figure 8 Schematically shows the hardware architecture diagram of the computer device according to Embodiment 3 of the present application. Detailed implementation manners

[0018] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0019] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between various embodiments may be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions appears to be contradictory or unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present application.

[0020] In the description of the present application, it should be understood that the numerical labels before the steps do not identify the order of execution of the steps, but are only used to facilitate the description of the present application and distinguish each step, and thus cannot be understood as a limitation to the present application.

[0021] First, the following provides the term explanations involved in the present application: List: A list form that ranks or classifies relevant objects according to certain rules and standards for display.

[0022] Natural Language Processing (NLP): An interdisciplinary field in computer science, artificial intelligence, and linguistics, mainly studying how to enable a computer to understand and process human natural language.

[0023] Secondly, to facilitate the understanding of the technical solutions provided by the embodiments of the present application by those skilled in the art, the related technologies are described below: Currently, when many live streaming platforms generate rankings, they generally rely on data such as the number of online viewers, bullet chat data, and viewing duration corresponding to the live content to generate the rankings. As a result, some live content with longer durations (such as movies and TV dramas) occupy the top positions in the rankings. However, these live content with high viewing durations are not necessarily the most popular ones. They only rank high in the rankings because of their long viewing durations. Other live content with higher interactivity is usually more popular but cannot rank high in the rankings due to its short live duration. Therefore, the current ranking generation method cannot accurately reflect the popularity of live content and has low accuracy.

[0024] For this reason, the embodiments of this application provide a ranking generation technical solution. In this technical solution, by classifying the live content and using different scoring factors for live rooms of different content types, the generated rankings can more reasonably reflect the popularity of live content and improve the accuracy of ranking generation. See the following for details.

[0025] Finally, for ease of understanding, an exemplary operating environment is provided below.

[0026] As Figure 1 shown, the environmental schematic diagram includes a service platform 2, a network 4, and a client 6, where: The service platform 2 can be composed of a single or multiple computing devices. These multiple computing devices can include virtualized computing instances. The virtualized computing instances can include virtual machines, such as emulations of computer systems, operating systems, servers, etc. The computing devices can load virtual machines based on virtual images and / or other data that define specific software (such as operating systems, dedicated applications, servers) for emulation. As the demand for different types of processing services changes, different virtual machines can be loaded and / or terminated on one or more computing devices. A hypervisor can be implemented to manage the use of different virtual machines on the same computing device.

[0027] The service platform 2 can be configured to communicate with the client 6, etc. through the network 4. The network 4 includes various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and / or the like. The network 4 can include physical links, such as coaxial cable links, twisted pair cable links, fiber optic links, and their combinations, or wireless links, such as cellular links, satellite links, Wi-Fi links, etc.

[0028] The service platform 2 can provide services such as storage, reading, writing, querying, deleting, etc., such as providing viewing services for live content to the client.

[0029] The client 6 can be an electronic device running operating systems such as Windows, Android™, or iOS, such as a smartphone, tablet device, laptop computer, virtual reality device, gaming device, set-top box, in-vehicle terminal, or smart TV. Based on the above operating systems, various application programs can be run, such as an application program for watching live broadcasts.

[0030] The client 6 can provide / configure a user access page for manipulating the service platform 2 or uploading an object, etc.

[0031] It should be noted that the above devices are exemplary, and in different scenarios or according to different requirements, the number and types of devices can be adjusted.

[0032] Next, taking the service platform as the execution subject, the technical solutions of the present application will be introduced through multiple embodiments. It should be noted that these embodiments can be implemented in various different forms and should not be construed as being limited only to the embodiments described herein.

[0033] Embodiment 1 Figure 2 A flowchart of a list generation method according to Embodiment 1 of the present application is schematically shown.

[0034] As Figure 2 shown, the list generation method may include steps S200 to S206, where: Step S200: Obtain the live interaction data of each live broadcast room and determine the content type of each live broadcast room.

[0035] Step S202: Obtain the scoring factors corresponding to each content type, where different content types correspond to different scoring factors, and the scoring factors include the weights of the live interaction data.

[0036] Step S204: Determine the score of each live broadcast room based on the live interaction data and the scoring factors.

[0037] Step S206: Generate a list based on the scores of each live broadcast room.

[0038] The list generation method provided in this embodiment obtains the live interaction data of each live broadcast room, determines the content type of each live broadcast room, obtains the scoring factors corresponding to each content type, determines the score of each live broadcast room based on the live interaction data and the scoring factors, and generates a list based on the scores of each live broadcast room. Since different content types correspond to different scoring factors, and the scoring factors include the weights of the live interaction data, the live broadcast rooms of different content types can be scored according to different scoring rules, so that the generated list can reasonably reflect the popularity of the live content and improve the accuracy of list generation.

[0039] The following is combined withFigure 2 , elaborate on each step in steps S200 to S206 and optional other steps in detail.

[0040] Step S200 , obtain the live interaction data of each live broadcast room and determine the content type of each live broadcast room.

[0041] The live interaction data may include data such as the number of likes, comments, shares, bullet screen data, viewing duration, gift data, return visit data, and voting and Q&A data. Specifically, the service platform can count the live interaction data of each live broadcast room and obtain the live interaction data of each live broadcast room in real time when a list needs to be generated.

[0042] The content type may include content such as movies, TV shows, games, popular science, and life sharing. In an optional embodiment, as Figure 3 shown, in step S200, determining the content type of each live broadcast room may include: Step S300, obtain the target data of each live broadcast room, where the target data includes at least partial data of the title, description, tags, and picture content of the live broadcast room.

[0043] Step S302, use a deep learning model and / or natural language processing method to determine the content type of each live broadcast room according to the target data.

[0044] The target data may include data such as the title of the live broadcast room, relevant description information of the live broadcast room, pre-tagged tags for the live broadcast room, and picture content; with the authorization of the live broadcast host in the live broadcast room, relevant data of the host can also be obtained as target data. Among them, the tags can be multiple preset tags provided by the host when creating the live broadcast room, and the host selects the corresponding tags according to the actual situation; while the picture content can be obtained by using machine learning to perform image or scene recognition on the picture content of the live broadcast room. For example, if fitness equipment and the host's fitness movements are identified in the picture, the picture content can be determined to be fitness-related content; if the picture content is a kitchen scene, the picture content can be determined to be food cooking-related content.

[0045] After obtaining the target data, these target data can be input into a pre-trained deep learning model to obtain the content type of the live broadcast room; or, a natural language processing method can be used to analyze the target data, and the content type of the live broadcast room can be determined according to the analysis results. For example, if the target data is the title of the live broadcast room, specifically "How to Make a Delicious Yu-Shiang Eggplant", then the title of the live broadcast room can be input into a pre-trained deep learning model, and the deep learning model identifies the content type of the live broadcast room as "Food Cooking" according to the title; another example, if the target data is the label of the live broadcast room, specifically "Honor of Kings", then a natural language processing method can be used to analyze the label to determine the content type of the live broadcast room as "Game Live Broadcast". When determining the content type of the live broadcast room, a deep learning model or a natural language processing method can be used alone to determine, or a deep learning model and a natural language processing method can be used in combination to determine.

[0046] In this embodiment, by obtaining the target data of each live broadcast room and using a deep learning model and / or a natural language processing method to determine the content type of each live broadcast room according to the target data, the content type of each live broadcast room can be accurately determined, thus facilitating the subsequent accurate scoring of each live broadcast room.

[0047] Step S202 , obtain the scoring factors corresponding to each content type, where different content types correspond to different scoring factors, and the scoring factors include the weights of live interaction data.

[0048] Specifically, the scoring factors corresponding to each content type can be pre-configured. After determining the content type of the live broadcast room, the scoring factors corresponding to each content type can be obtained according to this configuration. For example, a mapping table of content type and scoring factors can be configured. After determining the content type of a certain live broadcast room, the scoring factors corresponding to this live broadcast room can be obtained according to this content type. Taking the mapping table as an example, assuming that the live interaction data includes two items: the number of bullet screens and the viewing duration, the mapping table can be: for the content type of game live broadcast, the weight corresponding to the number of bullet screens can be 0.6, and the weight corresponding to the viewing duration can be 0.4; for the content type of TV drama, the weight corresponding to the number of bullet screens can be 0.4, and the weight corresponding to the viewing duration can be 0.6.

[0049] Step S204 , determine the score of each live broadcast room based on the live interaction data and the scoring factors.

[0050] Specifically, the live broadcast interaction data of each live broadcast room can be scored, and then the scores of the live broadcast interaction data are multiplied by the scoring factors obtained according to the content type of the live broadcast room to obtain the final scores of each live broadcast room. Continuing with the previous example, if the game live broadcast room A and the TV live broadcast room B are scored respectively, first score according to the live broadcast interaction data of the game live broadcast room A and the TV live broadcast room B. Suppose the results are: the score of the number of bullet screens in the game live broadcast room A is 80, and the score of the viewing duration is 60; the score of the number of bullet screens in the TV live broadcast room B is 50, and the score of the viewing duration is 80. Then the final score of the game live broadcast room A can be: 80 0.6 + 60 0.4 = 72, and the final score of the TV live broadcast room B can be: 50 0.4 + 80 0.6 = 68. Among them, when scoring the live broadcast interaction data, a unified scoring standard configured in advance can be used for scoring. For example, for the number of bullet screens, the scores corresponding to the number of bullet screens in three ranges can be specified. If it falls into the first range, the score is 80 - 100. If it falls into the second range, the score is 60 - 79. If it falls into the third range, the score is 0 - 59. Specifically, it can be set according to the actual situation, and no specific limitation is made here.

[0051] Step S206 , and generate a list based on the scores of each live broadcast room.

[0052] Specifically, in the case of obtaining the scores of each live broadcast room, the scores of each live broadcast room can be sorted in descending order, and a list can be generated according to the sorting results.

[0053] It can be understood that since the live broadcast interaction data may include some low-quality data, directly scoring based on the live broadcast interaction data may not accurately reflect the popularity of the live broadcast content. Therefore, the quality of the interaction can be further determined and then combined with the quality of the interaction for scoring. The following are some exemplary solutions: In an alternative embodiment, as Figure 4 shown, in step S204, determining the score of each live broadcast room based on the live broadcast interaction data and the scoring factor may include: Step S400, obtaining the target interaction data in the live broadcast interaction data.

[0054] Step S402, determining the interaction quality of each live broadcast room based on the target interaction data.

[0055] Step S404, determining the score of each live broadcast room based on the live broadcast interaction data, the scoring factor, and the interaction quality.

[0056] The target interaction data may include data that can evaluate the quality, such as barrage data or the number of comments. When determining the interaction quality of each live streaming room based on the target interaction data, it can be determined from aspects such as the proportion of low-quality content included in the target interaction data, the sentiment of the target interaction data, the preference for the live streaming room, and the depth of discussion. For example, if the proportion of low-quality content included in the target interaction quality is relatively high, it is determined that the interaction quality is low; for another example, if the sentiment of the target interaction data is relatively positive, it is determined that the interaction quality is high, etc. Among them, low-quality content can be some meaningless or low-value interaction content, such as multiple repeated comments sent by the same user. When determining the interaction quality, it can be obtained by analyzing the target interaction data through natural language processing methods and according to the analysis results.

[0057] When determining the interaction quality of each live streaming room, it can be based on the target interaction data to score the interaction quality of the live streaming room, and there is also a weight corresponding to the score of the interaction quality in the scoring factors; when determining the score of each live streaming room based on the live streaming interaction data, scoring factors, and interaction quality, it can be multiplying each live streaming interaction data by the corresponding weight, and at the same time multiplying the interaction quality by its corresponding weight, and finally adding them up to get the total score of the live streaming room. For example, if the live streaming interaction data includes the number of barrages and the viewing duration, the corresponding weights are 0.4 and 0.3, the scores are 80 and 60 respectively, and the weight corresponding to the interaction quality is 0.3 and the score is 50, then the score of the live streaming room is: 80 0.4 + 60 0.3 + 50 0.3 = 65. Optionally, the interaction quality can also be used as a scoring adjustment factor to adjust the score of the corresponding part in the live streaming room. For example, for live streaming room A, the interaction quality is determined to be 0.5 according to the target interaction data. The live streaming interaction data includes the number of barrages and the viewing duration, the corresponding weights are 0.6 and 0.4, and the scores are 80 and 60 respectively. The score of the live streaming room can be: (80 0.5) 0.6 + 60 0.4 = 48, where the interaction quality 0.5 is used as the scoring adjustment factor for the part of the number of barrages.

[0058] In this embodiment, by obtaining the target interaction data in the live streaming interaction data, determining the interaction quality of each live streaming room based on the target interaction data, and determining the score of each live streaming room based on the live streaming interaction data, scoring factors, and interaction quality, the score of the live streaming room can be combined with the interaction quality, so that the score of the live streaming room can more reasonably reflect the true popularity of the live streaming content and improve the accuracy of generating the list.

[0059] In an alternative embodiment, in step S402, based on the target interaction data, determine the interaction quality of each live streaming room, such asFigure 5 As shown, it may include: Step S500: Perform sentiment analysis and semantic depth analysis on the target interaction data using natural language processing methods to obtain a sentiment analysis result and a semantic depth analysis result.

[0060] Step S502: Determine the interaction quality of each live streaming room based on the sentiment analysis result and the semantic depth analysis result.

[0061] Specifically, the target interaction data can be preprocessed. For example, special or meaningless characters and stop words in the target interaction data can be removed; then, the target interaction data is tokenized using natural language processing, and the text in the target interaction data is decomposed into words or phrases, and then sentiment analysis and semantic depth analysis are performed based on the decomposed words or phrases to obtain a sentiment analysis result and a semantic depth analysis result. Among them, when performing sentiment analysis, it can be achieved by constructing a sentiment dictionary or using an existing sentiment dictionary. For example, the sentiment dictionary can include positive, negative, and neutral sentiments. The sentiment of the target interaction data can be classified through natural language processing to determine whether the target interaction data belongs to positive, negative, or neutral sentiments to obtain the sentiment analysis result. When the target interaction data contains multiple sentiments, the multiple sentiments can be synthesized to determine the most prominent sentiment as the sentiment analysis result. Optionally, a sentiment recognition model can also be trained through machine learning methods, and then the trained sentiment recognition model is used to obtain the sentiment analysis result of the target interaction data. When performing semantic depth analysis, named entity recognition can be used on the decomposed words or phrases to identify key entities in the target interaction data, such as person names, place names, products, roles, etc., and the semantic coherence, relevance, involved topics, and included levels of the relevant text are determined based on the context information and key entity information of the target interaction data to obtain the semantic depth analysis result of the target interaction data. Among them, the semantic depth analysis result can be, for example, deep, moderate, or shallow, etc.

[0062] The sentiment analysis result and the semantic depth analysis result can each correspond to a weight. After obtaining the sentiment analysis result and the semantic depth analysis result, scores can be respectively calculated based on the sentiment analysis result and the semantic depth analysis result, and then the scores of the two results are multiplied by the corresponding weights to determine the interaction quality of each live streaming room. For example, if the interaction quality is in the form of a score, and the sentiment analysis result and the semantic depth analysis result are also in the form of a score, the sentiment analysis result of a certain live streaming room B is 60 points, the semantic depth analysis result is 80 points, and the weights corresponding to the sentiment analysis result and the semantic depth analysis result are 0.5 and 0.5 respectively, then the interaction quality of live streaming room B can be: 60 × 0.5 + 80 × 0.5 = 70.

[0063] In this embodiment, by using natural language processing methods to perform sentiment analysis and semantic depth analysis on the target interaction data, sentiment analysis results and semantic depth analysis results are obtained. Based on the sentiment analysis results and semantic depth analysis results, the interaction quality of each live room is determined, which can effectively and reasonably determine the interaction quality of each live room through sentiment analysis and semantic depth analysis, thereby improving the accuracy of determining the score of each live room and the accuracy of list generation in combination with the interaction quality subsequently.

[0064] In an alternative embodiment, the content type includes a target content type, the duration of the target content type is greater than a preset threshold, and the scoring factor corresponding to the target content type further includes a duration penalty factor.

[0065] Among them, the preset threshold can be, for example, 2 hours. The target content type can specifically include content types with longer durations such as movies and TVs. For the target content type, in addition to the weight of the live interaction data, its scoring factor also includes a duration penalty factor. The duration penalty factor can be, for example, 0.5, 0.6 or 0.4, etc. In practical applications, the duration penalty factor can also be determined according to the duration. For live content with a longer duration, the duration penalty factor is smaller. For example, the duration penalty factor for a live duration between 2 hours and 4 hours can be 0.6, and the duration penalty factor for a live duration between 4 hours and 6 hours can be 0.5... Specifically, when determining the score of a live room of the target content type, the part of the live interaction data related to the duration (such as the viewing duration) needs to be multiplied by the duration penalty factor. For example, if the content type of a certain live room C is the target content type, its bullet screen quantity score and viewing duration score are 60 and 80 respectively, the corresponding weights are 0.4 and 0.6 respectively, and the duration penalty factor is 0.6, then the score of this live room C can be: 60 0.4 + (80 0.6) 0.6 = 52.8.

[0066] In this embodiment, for content types with a duration greater than the preset threshold, a duration penalty factor is also included, which can further adjust the proportion of the duration in the live room score, so that the score can more reasonably reflect the popularity of the live content and improve the accuracy of list generation.

[0067] In an alternative embodiment, the list generation method of the embodiment of the present application may further include: obtaining the historical interaction data of the current user under the authorization of the current user; correspondingly, in step S204, based on the live interaction data and the scoring factor, determining the score of each live room, as Figure 6 shown, may include: Step S600, adjusting the scoring factor based on the historical interaction data to obtain an adjusted scoring factor.

[0068] Step S602: Determine the score of each live streaming room based on the live streaming interaction data and the adjusted scoring factors.

[0069] The current user can be any user currently logged in to any client. Considering that when a user clicks on the list, they usually expect to select their favorite live streaming rooms from the list for viewing. Therefore, in this embodiment, adjusting the scoring factors and scores according to the historical interaction data of the current user can make the generated list more in line with the user's such needs.

[0070] Specifically, the service platform can record the live streaming interaction data of the current user with the authorization of the current user; when the user logs in or when a list needs to be generated, obtain the historical interaction data of the current user; then analyze the historical interaction data, adjust the scoring factors according to the results of the analysis of the historical interaction data to obtain the adjusted scoring factors; and then determine the scores of each live streaming room based on the real-time live streaming interaction data of all users and the adjusted scoring factors. For example, if it is found through the analysis of the historical interaction data of the current user that the current user is more fond of game live streaming, then the weight of the content type of game live streaming in terms of the number of bullet screens, etc. can be increased according to the preset rules, while the weight of this content type in terms of viewing duration, etc. can be decreased, so that the live streaming rooms of the content type of game live streaming are more likely to rank at the top of the list. Among them, when adjusting the scoring factors according to the historical interaction data, it is only necessary to make the adjusted scoring factors more conducive to generating a list that conforms to the preference tendency of the current user, and there is no specific limitation here.

[0071] In this embodiment, by obtaining the historical interaction data of the current user with the authorization of the current user, adjusting the scoring factors based on the historical interaction data, and determining the scores of the live streaming rooms according to the live streaming interaction data and the adjusted scoring factors, the generated list can meet the needs of the current users of each client, and improve the personalization of the generated list.

[0072] To make the present application easier to understand, the following provides an exemplary application.

[0073] 1. Assume that live streaming room A is a game live stream. The system generates a comprehensive score based on data such as the interaction quality, viewing duration, and return visit rate of the audience. Among them, the interaction quality corresponding to the game live stream type accounts for 40% of the score, the audience stickiness accounts for 30%, and the viewing duration accounts for 30%. The interaction quality is determined according to the emotional tendency and discussion depth (semantic depth) of the bullet screens, and the viewing stickiness is determined according to live streaming interaction data such as the return visit rate and the average number of views.

[0074] The scores are as follows: Interaction quality score: 85, audience stickiness score: 80, and viewing duration score: 60; Then the comprehensive score is calculated as: Stotal = 85 0.4 + 80 0.3 + (60 1) 0.3 = 34 + 24 + 18 = 76。

[0075] Among them, (60 1)indicates that the duration penalty factor is 1, which means there is no corresponding penalty for the viewing duration.

[0076] 2. Assume that live broadcast room B is a movie live broadcast, with a relatively long viewing duration of the audience but less interaction. The system generates a comprehensive score based on data such as the interaction quality, viewing duration, and return visit rate of the audience. Among them, the interaction quality corresponding to the movie live broadcast type accounts for 30% of the score, the audience stickiness accounts for 30%, and the viewing duration accounts for 40%. At the same time, the system's duration penalty factor for the movie live broadcast type is 0.5, and the score of the viewing duration will be multiplied by the penalty factor.

[0077] Assume the scores are as follows: The interaction quality score is 60, the audience stickiness score is 80, and the viewing duration score is 90; Then the comprehensive score is calculated as follows: Stotal = 60 0.3 + 80 0.3 + (90 0.5) 0.4 = 18 + 24 + 18 = 60 In this exemplary application, the viewing duration of live broadcast room A is short, but the interaction quality is high, and the final ranking is higher than that of live broadcast room B; while live broadcast room B, although having a long viewing duration, has a low interaction quality and a duration penalty factor, so the final ranking is lower than that of live broadcast room A. It can be seen that different weights are used for game live broadcasts and movie live broadcasts, and at the same time, the duration penalty factor is introduced, which can make the rankings of live broadcast types with high interaction quality, such as game live broadcasts, move forward, reduce the proportion of the duration in the generation of the list, so that the generated list can more reasonably reflect the true popularity of the live broadcast room and improve the accuracy of the list generation.

[0078] Embodiment 2 Figure 7 Schematically shows a block diagram of a list generation device according to Embodiment 2 of the present application. The device can be divided into one or more program modules. One or more program modules are stored in a storage medium and executed by one or more processors to complete the embodiments of the present application. The program modules referred to in the embodiments of the present application refer to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment. As Figure 7As shown, the device 700 may include: a determination module 710, an acquisition module 720, a scoring module 730, and a generation module 740, where: The determination module 710 is configured to obtain the live interaction data of each live broadcast room and determine the content type of each said live broadcast room; The acquisition module 720 is configured to obtain the scoring factors corresponding to each said content type, where different content types correspond to different scoring factors, and the scoring factors include the weights of the live interaction data; The scoring module 730 is configured to determine the score of each said live broadcast room based on the live interaction data and the scoring factors; The generation module 740 is configured to generate a leaderboard based on the score of each said live broadcast room.

[0079] In an alternative embodiment, the determination module 710 is further configured to: Obtain the target data of each said live broadcast room, where the target data includes at least partial data of the title, description, tags, and video content of the live broadcast room; Use a deep learning model and / or natural language processing method to determine the content type of each said live broadcast room according to the target data.

[0080] In an alternative embodiment, the scoring module 730 is further configured to: Obtain the target interaction data in the live interaction data; Determine the interaction quality of each said live broadcast room based on the target interaction data; Determine the score of each said live broadcast room based on the live interaction data, the scoring factors, and the interaction quality.

[0081] In an alternative embodiment, the scoring module 730 is further configured to: Use a natural language processing method to perform sentiment analysis and semantic depth analysis on the target interaction data to obtain a sentiment analysis result and a semantic depth analysis result; Determine the interaction quality of each said live broadcast room based on the sentiment analysis result and the semantic depth analysis result.

[0082] In an alternative embodiment, the content type includes a target content type, the duration of the target content type is greater than a preset threshold, and the scoring factors corresponding to the target content type further include a duration penalty factor.

[0083] In an alternative embodiment, the device 700 is further configured to: Obtain the historical interaction data of the current user when the current user authorizes; Correspondingly, the scoring module 730 is further configured to: Adjust the scoring factor based on the historical interaction data to obtain an adjusted scoring factor; Determine the score of each live broadcast room based on the live interaction data and the adjusted scoring factor.

[0084] Embodiment III Figure 8 Schematically shows a hardware architecture diagram of a computer device 10000 suitable for implementing the list generation method according to Embodiment III of the present application. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a wearable device, a tablet computer, a personal computer, a vehicle-mounted terminal, a game console, a virtual device, a workbench, a digital assistant, a set-top box, a robot, etc. In other embodiments, the computer device 10000 may be a rack-mounted server, a blade server, a tower server or a cabinet server (including an independent server or a server cluster composed of multiple servers). As Figure 8 shown, the computer device 10000 includes, but is not limited to: a memory 10010, a processor 10020, and a network interface 10030 that can be communicatively linked to each other through a system bus. Among them: The memory 10010 includes at least one type of computer-readable storage medium. The readable storage medium includes flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10010 may be an internal storage module of the computer device 10000, such as the hard disk or memory of the computer device 10000. In other embodiments, the memory 10010 may also be an external storage device of the computer device 10000, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 10000. Of course, the memory 10010 may also include both the internal storage module and the external storage device of the computer device 10000. In this embodiment, the memory 10010 is generally used to store the operating system and various application software installed on the computer device 10000, such as the program code of the list generation method. In addition, the memory 10010 may also be used to temporarily store various data that have been output or will be output.

[0085] The processor 10020 can be a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other chips in some embodiments. The processor 10020 is generally used to control the overall operation of the computer device 10000, such as performing control and processing related to data interaction or communication with the computer device 10000. In this embodiment, the processor 10020 is used to run the program code stored in the memory 10010 or process data.

[0086] The network interface 10030 may include a wireless network interface or a wired network interface. The network interface 10030 is generally used to establish a communication link between the computer device 10000 and other computer devices. For example, the network interface 10030 is used to connect the computer device 10000 to an external terminal through a network, and establish a data transmission channel and a communication link between the computer device 10000 and the external terminal. The network can be an enterprise intranet (Intranet), the Internet, the Global System of Mobile communication (abbreviated as GSM), Wideband Code Division Multiple Access (abbreviated as WCDMA), 4G network, 5G network, Bluetooth, Wi-Fi and other wireless or wired networks.

[0087] It should be noted that Figure 8 Only the computer device with components 10010 - 10030 is shown, but it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented.

[0088] In this embodiment, the list generation method stored in the memory 10010 can also be divided into one or more program modules and executed by one or more processors (such as the processor 10020) to complete the embodiments of the present application.

[0089] Embodiment 4 The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the list generation method in the embodiments are implemented.

[0090] In this embodiment, the computer-readable storage medium includes flash memory, hard disks, multimedia cards, card-type memories (such as SD or DX memories, etc.), random access memories (RAM), static random access memories (SRAM), read-only memories (ROM), electrically erasable programmable read-only memories (EEPROM), programmable read-only memories (PROM), magnetic memories, magnetic disks, optical disks, etc. In some embodiments, the computer-readable storage medium may be an internal storage unit of a computer device, such as the hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc., equipped on the computer device. Of course, the computer-readable storage medium may also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the computer-readable storage medium is generally used to store the operating system installed on the computer device and various application software, such as the program code of the list generation method in the embodiment. In addition, the computer-readable storage medium may also be used to temporarily store various data that have been output or will be output.

[0091] Embodiment 5 The embodiment of the present application further provides a computer program product, including a computer program, which when executed by a processor implements the method in the above embodiment.

[0092] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the embodiments of the present application can be implemented by a general computer device. They can be concentrated on a single computer device or distributed on a network composed of multiple computer devices. Optionally, they can be implemented by program codes executable by the computer device. Thus, they can be stored in a storage device and executed by the computer device. And in some cases, the steps shown or described can be executed in a different order from here, or they can be separately made into individual integrated circuit modules, or multiple modules or steps among them can be made into a single integrated circuit module to implement. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0093] It should be noted that the above are only the preferred embodiments of the present application, and do not limit the patent protection scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A list generation method, characterized in that, The method includes: Obtaining the live interaction data of each live broadcast room and determining the content type of each live broadcast room; Obtaining a scoring factor corresponding to each content type, where different content types correspond to different scoring factors, and the scoring factor includes the weight of the live interaction data; Determining the score of each live broadcast room based on the live interaction data and the scoring factor; Generating a list based on the scores of each live broadcast room.

2. The method according to claim 1, wherein The determining the content type of each live broadcast room includes: Obtaining the target data of each live broadcast room, where the target data includes at least partial data of the title, description, tags, and video content of the live broadcast room; Using a deep learning model and / or natural language processing method to determine the content type of each live broadcast room according to the target data.

3. The method according to claim 1, wherein The determining the score of each live broadcast room based on the live interaction data and the scoring factor includes: Obtaining the target interaction data in the live interaction data; Determining the interaction quality of each live broadcast room based on the target interaction data; Determining the score of each live broadcast room based on the live interaction data, the scoring factor, and the interaction quality.

4. The method according to claim 3, wherein The determining the interaction quality of each live broadcast room based on the target interaction data includes: Using a natural language processing method to perform sentiment analysis and semantic depth analysis on the target interaction data to obtain a sentiment analysis result and a semantic depth analysis result; Determining the interaction quality of each live broadcast room based on the sentiment analysis result and the semantic depth analysis result.

5. The method according to claim 1, characterized in that, The content type includes a target content type, the duration of the target content type is greater than a preset threshold, and the scoring factor corresponding to the target content type further includes a duration penalty factor.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Obtaining the historical interaction data of the current user under the authorization of the current user; Correspondingly, the determining the score of each live broadcast room based on the live interaction data and the scoring factor includes: Adjusting the scoring factor based on the historical interaction data to obtain an adjusted scoring factor; Determining the score of each live broadcast room based on the live interaction data and the adjusted scoring factor.

7. A list generation device, characterized in that, The device includes: A determination module, configured to obtain the live interaction data of each live broadcast room and determine the content type of each live broadcast room; An acquisition module, configured to obtain a scoring factor corresponding to each content type, where different content types correspond to different scoring factors, and the scoring factor includes the weight of the live interaction data; A scoring module, configured to determine the score of each live broadcast room based on the live interaction data and the scoring factor; A generation module, configured to generate a list based on the scores of each live broadcast room.

8. A computer device, characterized in that, Includes: At least one processor; And A memory communicatively connected to the at least one processor; wherein: The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method described in any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method described in any one of claims 1 to 6 are implemented.

Images