A live room information display method, system, device and medium

Abstract

The application relates to the technical field of live content display and interaction optimization, and discloses a live room information display method, system, device and medium, which comprises the following steps: collecting audience data and analyzing audience viewing behavior in real time; automatically matching viewing modes according to the audience viewing behavior and displaying different information levels for different viewing modes; analyzing the audience's barrage and automatically screening the barrage priority; and dynamically displaying live content according to the adjusted information level and barrage priority. Through intelligent dynamic adjustment of the display level of live content and the barrage priority, more personalized and efficient information display is realized. According to the viewing behavior, interaction frequency and device network status of the audience, the system automatically adjusts the information density and display level, avoids information overload or loss, and improves the user experience. Meanwhile, based on the emotional analysis of the barrage, the interaction frequency and the content relevance, high-value barrages are preferentially displayed, and the audience's interaction sense and participation are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of live content display and interaction optimization, in particular to a live room information display method, system, device and medium. BACKGROUND

[0002] With the popularity of the Internet and mobile devices, the live industry has developed rapidly and has become an important part of people's daily life. Whether in social media platforms, online education, e-commerce, or in entertainment live streaming, game live streaming and other fields, real-time interactive live content occupies an important position. The quality of live content and the interactive experience of the audience directly affect the activity and user stickiness of the platform, therefore, how to efficiently display live content and improve the interactive experience of the audience has become a problem to be solved for major live platforms.

[0003] Traditional live content display methods have many problems. First, current live information display is usually fixed, when the amount of information is too much, the audience will feel that the interface is messy and it is difficult to quickly find key information; when the information is too little, the audience may feel that the content is empty and cannot meet their needs. Second, the live platform's barrage information has large traffic and strong real-time, but many barrages lack a screening mechanism, causing valuable barrage information to be easily submerged, affecting the interactive effect. In addition, the viewing behavior, interactive frequency, device environment and other factors of the audience are different, the existing display mode cannot dynamically adjust according to these factors, resulting in unsatisfactory information display effect.

[0004] In order to improve the efficiency of live content display, more and more research has begun to focus on how to intelligently adjust the information density and content level based on the behavior and environment data of the audience. For example, how to adjust the amount of information displayed in real time for different audience groups and different viewing modes to avoid information overload or information deficiency has become a difficult problem to be solved. In addition, how to dynamically adjust the priority of the barrage based on the interactive situation of the audience (such as the emotional tendency of the barrage, the interactive frequency, etc.) to more effectively highlight valuable interactive information is also an important research direction.

[0005] The current live platform still lacks a system that can intelligently adjust the information display level and barrage priority, and cannot achieve accurate content optimization according to the behavior, emotion and interaction of the audience. Therefore, it is urgent to develop an intelligent dynamic display method based on audience behavior analysis, emotion recognition and interactive data to improve the display efficiency of live content, optimize the user interactive experience, and improve the audience stickiness and satisfaction of the platform. SUMMARY

[0006] In view of the above problems, the present application is proposed.

[0007] Therefore, the present application solves the technical problem of how to intelligently and dynamically adjust the display level of live content and the priority of barrage based on audience behavior, interaction data and device network conditions, so as to optimize live information display and interactive experience.

[0008] To solve the above technical problems, the present application provides the following technical solutions: a live room information display method, comprising:

[0009] Collecting audience data and analyzing audience viewing behavior in real time;

[0010] According to the audience viewing behavior, automatically matching the viewing mode and displaying different information levels for different viewing modes;

[0011] Analyzing the audience's barrage and automatically filtering the barrage priority;

[0012] According to the adjusted information level and barrage priority, dynamically displaying live content.

[0013] As a preferred scheme of the live room information display method of the present application, wherein: the audience data includes viewing behavior data, interaction behavior data, emotional and emotional data, and device and network data;

[0014] The real-time analysis of audience viewing behavior includes extracting behavior characteristics from audience data;

[0015] The behavior characteristics include user activity characteristics, viewing mode characteristics, emotional tendency characteristics, user type characteristics, participation characteristics and device and network characteristics;

[0016] The user activity characteristics include audience online duration and interaction frequency;

[0017] The viewing mode characteristics include dwell time and browsing speed;

[0018] The emotional tendency characteristics include emotional score and user type characteristics;

[0019] The participation characteristics include interaction depth and interaction breadth;

[0020] The device and network characteristics include device type and network condition;

[0021] Defining the viewing mode based on the behavior characteristics;

[0022] The viewing mode includes fast browsing mode, long dwell time mode, high interaction mode, low interaction mode and device and network optimization mode.

[0023] As a preferred embodiment of the live room information display method of the present invention, the step of displaying different information levels for different viewing modes includes dividing multiple audience data into multiple layers based on the viewing mode category, with each layer corresponding to a viewing mode category.

[0024] Based on the distribution ratio of various viewing modes in the overall audience data, the sampling ratio of each layer is determined, and samples are randomly drawn from each layer as the training set. The formula is as follows:

[0025] ;

[0026] ;

[0027] in, Indicates the first The sampling ratio of a layer is equal to the proportion of the audience data from the m-th layer in the total audience data. Indicates the first The number of samples in the audience data. This represents the total number of samples in the overall audience data. Represents the final training set, RS Indicates from the first layer In proportion A randomly selected sample set, where RS represents random sampling. Hierarchical structure of audience data;

[0028] The similarity between behavioral characteristics of viewers is calculated using Euclidean distance, expressed by the formula: ;

[0029] in, This represents the weighted Euclidean distance. Indicates the first Behavioral characteristics of the audience Indicates the first The audience member at the The values ​​that can be taken on each behavioral feature Indicates the first Behavioral characteristics of the audience Indicates the first The audience member at the The values ​​that can be taken on each behavioral feature Indicates the first The weight of each behavioral feature, The total dimensions representing behavioral characteristics;

[0030] choose Calculate the similarity between the behavioral characteristics of each viewer and other viewers, and find the viewer most similar to the current viewer. The approximate KNN model is constructed, and the formula is as follows:

[0031] ;

[0032] wherein, indicates a set of neighbors most similar to the audience ; indicates that the behavior characteristics of the audience are selected, indicates that the behavior characteristics of the audience are selected, indicates the weighted Euclidean distance between the audience and the audience ; indicates that the behavior characteristics are selected from all audiences in the training set ;

[0033] The approximate KNN model is constructed, and the approximate KNN model is trained based on the training set after hierarchical sampling, by using a local sensitive hash data index structure to preprocess the training set.

[0034] For the audience newly entering the live room, the behavior characteristics of the audience are extracted in real time, and the viewing mode of the new audience is determined by majority voting according to the viewing mode categories of the nearest neighbors by using the approximate KNN model, and the formula is as follows: ;

[0035] wherein, indicates the predicted viewing mode of the new audience, MV indicates majority voting, indicates the viewing mode category of the audience ; indicates that the viewing modes of the nearest neighbors of the new audience are selected to vote.

[0036] As a preferred scheme of the live room information display method, the different information levels are displayed according to different viewing modes, including automatically switching to a basic information level to display the most core basic information and reducing interface elements when the system detects that the audience is in a fast browsing mode.

[0037] When the audience stays on the live page for a long time, the system automatically switches to an intermediate information level to display more interactive data and discussion topics.

[0038] When the system detects that the audience is in a high-interactive mode, the system switches to a deep information level to display detailed interactive information and content details.

[0039] ​When the system detects that the audience is in a low interaction mode, the system maintains a medium level of information, providing necessary interaction information, and encourages the audience to increase interaction;

[0040] When the system detects that the audience is in a device and network optimization mode, in a low bandwidth environment, the display of high-resolution images and videos is reduced.

[0041] As a preferred scheme of the live room information display method, wherein: the analysis of the audience's barrage includes emotional analysis, interaction frequency calculation and content relevance analysis;

[0042] The emotional analysis includes using natural language processing technology to analyze the emotions of all audience barrages; the emotional score of each barrage is divided into positive, negative and neutral categories;

[0043] The interaction frequency calculation includes counting the interaction frequency of each barrage, and the barrage with high interaction frequency indicates that the barrage has higher interaction value;

[0044] The interaction frequency includes likes, forwards and replies;

[0045] The content relevance analysis includes judging the relevance of each barrage to the current live content through semantic analysis technology.

[0046] As a preferred scheme of the live room information display method, wherein: the automatic screening of barrage priority includes designing a comprehensive scoring formula, which comprehensively considers emotional analysis, interaction frequency and content relevance, and the formula is expressed as:

[0047]

[0048] Wherein, represents the comprehensive score, represents the attenuation factor, represents the time variable, represents the barrage at time Emotional score; represents the interaction frequency of the barrage at time ; represents the relevance score of the barrage to the current live content theme ;

[0049] The value range of the comprehensive score is ;

[0050] When ​When the value of the first parameter is greater than the value of the second parameter, it indicates that the barrage has high interaction value, positive emotion and high relevance with the live content, and is high-priority barrage.

[0051] When the value of the first parameter is less than the value of the second parameter, it indicates that the barrage has low interaction value, negative emotion and low relevance with the live content, and is low-priority barrage.

[0052] As a preferred scheme of the live room information display method, the dynamic display of the live content according to the adjusted information level and barrage priority comprises continuously monitoring the behavior change of the audience and the comprehensive score of the barrage, so as to ensure that the information display is synchronized with the current state of the audience.

[0053] When the viewing mode of the audience changes, the information level is automatically switched to the corresponding information level.

[0054] When the comprehensive score of the barrage changes, the display priority of the barrage is automatically adjusted in real time.

[0055] A live room information display system comprises:

[0056] A data collection module collects audience data and analyzes the viewing behavior of the audience in real time.

[0057] An information level module automatically matches the viewing mode according to the viewing behavior of the audience, and displays different information levels for different viewing modes.

[0058] A barrage analysis module analyzes the barrage of the audience and automatically filters the barrage priority.

[0059] A continuous adjustment module dynamically displays the live content according to the adjusted information level and barrage priority.

[0060] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method according to any one of the present application when executing the computer program.

[0061] A computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the method according to any one of the present application.

[0062] The live room information display method provided by the present application can realize more personalized and efficient information display by intelligently and dynamically adjusting the display level of the live content and the priority of the barrage. According to the viewing behavior, interaction frequency and device network status of the audience, the system automatically adjusts the information density and display level, avoids information overload or lack, and improves the user experience. At the same time, based on the emotional analysis, interaction frequency and content relevance of the barrage, high-value barrages are preferentially displayed, and the interaction and participation of the audience are enhanced. BRIEF DESCRIPTION OF DRAWINGS​

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0064] Figure 1 The overall flowchart of a live room information display method provided for the first embodiment of the present application. DETAILED DESCRIPTION

[0065] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings in the specification. Obviously, the described embodiments are only some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort should be within the protection scope of the present application.

[0066] Embodiment 1, refer to Figure 1 For an embodiment of the present application, a live room information display method is provided, which comprises:

[0067] S1: Collecting audience data to analyze audience viewing behavior in real time.

[0068] The audience data includes viewing behavior data, interactive behavior data, emotional and emotional data, and device and network data.

[0069] The real-time analysis of the audience viewing behavior comprises extracting behavior features from the audience data.

[0070] The behavior features include user activity features, viewing mode features, emotional tendency features, user type features, participation features and device and network features.

[0071] The user activity features include audience online duration and interaction frequency.

[0072] The viewing mode features include dwell time and browsing speed.

[0073] The emotional tendency features include emotional score and user type features.

[0074] The participation features include interaction depth and interaction breadth.

[0075] The device and network features include device type and network status.

[0076] The viewing mode is defined based on the behavior features.

[0077] The viewing modes include fast browsing mode, long-stay mode, high interaction mode, low interaction mode, and device and network optimization mode.

[0078] Furthermore, by comprehensively collecting and analyzing multidimensional audience data in real time, various behavioral characteristics can be extracted to accurately identify audience viewing patterns. Analyzing audience activity, viewing behavior, emotional tendencies, engagement, and device and network conditions allows for a better understanding of audience needs and preferences, providing a basis for subsequent intelligent content display and interactive optimization. Defining and identifying different viewing patterns, such as fast browsing, long-stay viewing, and high-interaction modes, helps to dynamically adjust the information display hierarchy and bullet screen priority, achieving a personalized live streaming experience and enhancing audience engagement and satisfaction.

[0079] S2: Automatically match the viewing mode based on the viewer's viewing behavior, and display different information levels for different viewing modes.

[0080] The method of displaying different information levels for different viewing modes includes dividing multiple audience data into multiple layers based on the viewing mode category, with each layer corresponding to a viewing mode category.

[0081] Based on the distribution ratio of various viewing modes in the overall audience data, the sampling ratio of each layer is determined, and samples are randomly drawn from each layer as the training set. The formula is as follows:

[0082] ;

[0083] ;

[0084] in, Indicates the first The sampling ratio of a layer is equal to the proportion of the audience data from the m-th layer in the total audience data. Indicates the first The number of samples in the audience data. This represents the total number of samples in the overall audience data. Represents the final training set, RS Indicates from the first layer In proportion A randomly selected sample set, where RS represents random sampling. The hierarchy of audience data.

[0085] The similarity between behavioral characteristics of viewers is calculated using Euclidean distance, expressed by the formula: ;

[0086] in, represents a weighted Euclidean distance, represents the behavior characteristics of the th viewer, represents the value of the th behavior characteristic of the th viewer, represents the behavior characteristics of the th viewer, represents the value of the th behavior characteristic of the th viewer, represents the weight of the th behavior characteristic, represents the total dimension of the behavior characteristics.

[0087] Selecting , calculating the similarity between the behavior characteristics of each viewer and other viewers, finding the most similar viewers to the current viewer, constructing an approximate KNN model, and the formula is represented as:

[0088] ;

[0089] wherein, represents the set of most similar neighbors of the viewer , represents selecting neighbors most similar to the behavior characteristics of the viewer , represents the weighted Euclidean distance between the th viewer and the th viewer; represents the value of the th behavior characteristic of the th viewer. represents selecting from the behavior characteristics of all viewers in the training set

[0090] Constructing an approximate KNN model includes preprocessing the training set using a local sensitive hash data index structure, training an approximate KNN model based on the training set after stratified sampling.

[0091] For a new viewer entering the live room, the behavior characteristics of the new viewer are extracted in real time, and the viewing mode of the new viewer is determined by majority voting according to the viewing mode category of the nearest neighbors through the approximate KNN model, and the formula is represented as: ;

[0092] wherein, represents the predicted viewing mode of the new viewer, MV represents majority voting, represents the viewing mode category of the th viewer. representing the watching mode of the new audience from the selecting their watching mode from the voting.

[0093] Further, by analyzing the behavioral characteristics of the audience, the system intelligently matches each new audience entering the live room with an appropriate watching mode, thereby displaying the most suitable information hierarchy for them. Through hierarchical sampling and similarity calculation based on Euclidean distance, the system can accurately identify other audiences similar to the new audience's behavior and vote according to their watching mode to predict the new audience's watching mode category. This method dynamically adjusts the density and content hierarchy of information display to meet the needs of different audiences, providing personalized and optimized user experience, avoiding information overload or scarcity, and thus improving audience interaction participation and satisfaction.

[0094] S3: Analyze the audience's barrage and automatically filter the barrage priority.

[0095] The analysis of the audience's barrage includes sentiment analysis, interaction frequency calculation, and content relevance analysis.

[0096] The sentiment analysis includes using natural language processing technology to analyze the sentiment of all audience barrages; the sentiment score of each barrage is divided into positive, negative, and neutral categories.

[0097] The interaction frequency calculation includes counting the interaction frequency of each barrage, and barrages with high interaction frequency represent higher interaction value.

[0098] The interaction frequency includes likes, forwards, and replies.

[0099] The content relevance analysis includes using semantic analysis technology to determine the relevance of each barrage to the current live content.

[0100] The sentiment analysis includes using natural language processing technology to analyze the sentiment of all audience barrages; the sentiment score of each barrage is divided into positive, negative, and neutral categories.

[0101] The formula of the sentiment analysis is:

[0102] ;

[0103] wherein, represents the sentiment score of the th barrage, represents the weight of the th barrage.

[0104] The interaction frequency calculation includes counting the interaction frequency of each bullet screen, and the bullet screen with high interaction frequency represents higher interaction value.

[0105] The formula of the interaction frequency calculation is:

[0106] ;

[0107] Wherein, represents the interaction frequency, is the number of likes of the th bullet screen, is the number of forwarding of the th bullet screen, is the number of replies of the th bullet screen, is the matching degree of the th bullet screen and the current live content, which is calculated by cosine similarity or deep learning model.

[0108] The interaction frequency includes likes, forwarding and replies.

[0109] The content relevance analysis includes judging the relevance of each bullet screen and the current live content through semantic analysis technology.

[0110] The formula of the content relevance analysis is:

[0111] ;

[0112] Wherein, represents the theme vector of the current live content, represents the theme vector of the th bullet screen, is the cosine similarity, represents the normalization factor of the th bullet screen, and the value range is [0, 1]; and are the vector lengths of live content and bullet screen respectively.

[0113] The automatic screening of bullet screen priority includes designing a comprehensive scoring formula, which comprehensively considers sentiment analysis, interaction frequency and content relevance, and the formula is:

[0114] ;

[0115] Wherein, represents the comprehensive score, represents the attenuation factor, represents the time variable, represents the sentiment score of the bullet screen at time ; represents the barrage in time interaction frequency; represents the barrage correlation score with the current live content theme .

[0116] the value range of the comprehensive score . .

[0117] when , it means that the barrage has high interaction value, positive emotion and high relevance to the live content, and is a high-priority barrage.

[0118] when , it means that the barrage has low interaction value, negative emotion and low relevance to the live content, and is a low-priority barrage.

[0119] Further, by comprehensively analyzing the emotional tendency, interaction frequency and content relevance of the barrage, the barrage with high interaction value is intelligently screened and preferentially displayed, thereby improving the interaction experience of the audience and the accuracy of the live content. Emotional analysis can identify the emotional attitude of the barrage, interaction frequency calculation reflects the popularity of the barrage, and content relevance analysis ensures that the barrage is highly matched with the current live content. Through the comprehensive scoring method, the priority of the barrage can be dynamically adjusted to avoid the interference of low-value or irrelevant barrages, so that the audience can quickly see meaningful interaction information and enhance the attractiveness and participation of the live broadcast.

[0120] S4: dynamically display the live content according to the adjusted information level and barrage priority.

[0121] According to the adjusted information level and barrage priority, dynamically displaying the live content includes continuously monitoring the behavior changes of the audience and the comprehensive score of the barrage to ensure that the information display is synchronized with the current state of the audience.

[0122] When the viewing mode of the audience changes, automatically switch to the corresponding information level.

[0123] When the comprehensive score of the barrage changes, real-time adjust the display priority of the barrage.

[0124] Further, by real-time monitoring and dynamically adjusting the audience behavior and barrage score, it is ensured that the live content display is always consistent with the needs and interaction state of the audience. The viewing mode of the audience and the interaction feedback of the barrage will directly affect their experience, and automatically switching the information level and adjusting the barrage priority can achieve personalized and accurate content presentation, avoid information overload or omission, improve the audience's sense of participation and interactivity, and thereby enhance the user experience and platform stickiness.

[0125] Embodiment 2, which is an embodiment of the present application, provides a live room information display system, comprising:

[0126] A data collection module analyzes audience viewing behavior in real time by collecting audience data.

[0127] An information level module automatically matches viewing modes according to audience viewing behavior and displays different information levels for different viewing modes.

[0128] A barrage analysis module analyzes audience barrages and automatically filters barrage priorities.

[0129] A continuous adjustment module dynamically displays live content according to adjusted information levels and barrage priorities.

[0130] Embodiment 3, which is an embodiment of the present application, is different from the previous two embodiments in that:

[0131] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application or the parts that essentially contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0132] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be specifically embodied in any computer-readable medium for use by an instruction execution system, device or equipment (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or equipment and execute the instructions) or in conjunction with these instructions execution systems, devices or equipment. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by an instruction execution system, device or equipment or in conjunction with these instruction execution systems, devices or equipment.

[0133] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer readable medium can be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example via an optical scanner, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, to generate an electronically readable version of the program, which can then be stored in the computer memory.

[0134] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and as in another embodiment, can be implemented using any or a combination of the following technologies, which are well known in the art: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0135] Embodiment 4, as an embodiment of the present application, provides a live room information display method, system, device and medium, in order to verify the beneficial effects of the present application, scientific demonstration is carried out through simulation experiment.

[0136] The experiment selects some live rooms of a live platform, collects real-time audience data and implements corresponding behavior analysis model, and dynamically displays and adjusts the live content.

[0137] In the implementation process, first, the audience's viewing behavior data is collected, including viewing time, interaction frequency (such as comments, likes, sharing, etc.), emotional tendency (emotional analysis of the bullet screen through natural language processing), device type and network condition, etc. All these data are transmitted to the central processing system in real time through API for real-time analysis and display adjustment.

[0138] Using hierarchical sampling method, the data set is layered according to the audience's viewing mode category (such as fast browsing mode, long time staying mode, high interaction mode, low interaction mode, etc.). By calculating the proportion of audience data in each layer, the sampling proportion of each layer is determined, and data is randomly extracted from each layer as the training set.

[0139] For each audience, the system extracts its behavioral characteristics, including dwell time, browsing speed, interaction frequency, etc., and calculates the similarity between audiences based on Euclidean distance through the approximate KNN algorithm, predicting the viewing mode of the audience.

[0140] The sentiment score of each bullet screen is analyzed using sentiment analysis technology, and the bullet screen is comprehensively scored by combining the interaction frequency and content relevance. Based on the score, the bullet screen with positive sentiment, high interaction frequency and related to the current live content is preferentially displayed.

[0141] According to the behavior mode of the audience, the information level is automatically adjusted. When the system identifies that the audience is in fast browsing mode, the simplified basic information is automatically displayed; when the audience stays for a long time, more interactive data and discussion topics are displayed; when the interaction frequency of the audience increases, more interactive information and content details are displayed; if the system detects that the audience is in low bandwidth mode, the resolution of images and videos is automatically reduced to ensure that the information display is not affected.

[0142] By monitoring the changes in audience behavior data in real time, the system automatically adjusts the content of information display and the priority of bullet screen according to the current viewing mode and sentiment analysis results of the audience, ensuring that the displayed content meets the current needs of the audience and improving user experience.

[0143] In order to verify the effect of the technical scheme, the following experiments were carried out:

[0144] A live broadcast platform was selected, and the viewing behavior data of 1000 different audiences of the platform was selected as the experimental sample. The experiment lasted for 48 hours, and the data collection included viewing behavior, interaction frequency, device type, network status, etc.

[0145] The system extracts and classifies five viewing modes through real-time monitoring, including fast browsing mode, long-time staying mode, high interaction mode, low interaction mode, and device and network optimization mode. The audience data sampling of each mode is as follows:

[0146] Fast browsing mode: audience proportion 24.5%;

[0147] Long-time staying mode: audience proportion 33.2%;

[0148] High interaction mode: audience proportion 18.6%;

[0149] Low interaction mode: audience proportion 14.7%;

[0150] Device and network optimization mode: audience proportion 9%.

[0151] Bullet screen priority analysis: a total of 120,000 bullet screen data were collected in the experiment. After sentiment analysis, the sentiment distribution of the bullet screen was:

[0152] Positive emotions: 45.3%;

[0153] Neutral emotions: 38.9%;

[0154] Negative emotions: 15.8%.

[0155] According to the interaction frequency and content relevance analysis, the comprehensive priority distribution of the bullet screen is as follows:

[0156] High priority bullet screen: 28.7%;

[0157] Medium priority bullet screen: 45.2%;

[0158] Low priority bullet screen: 26.1%.

[0159] Information level adjustment: In the implementation process, the display level of live content automatically switches in different modes. When the audience is in the long-time stay mode, the interactive data displayed increases by 32.5%; in the high-interactive mode, the content depth displayed increases by 29.8%; in the fast-browsing mode, the information is simplified by 18.6%.

[0160] From the experimental data analysis results, the live room information display method has a significant optimization effect, and has obvious innovation and advantage in user experience compared with the prior art.

[0161] Firstly, based on the hierarchical sampling of viewing modes and real-time analysis of audience behavior characteristics, the information display can be dynamically adjusted according to the actual needs of the audience. In traditional live platforms, information display often uses fixed layout and display form, which lacks flexibility, resulting in the problem of information overload or insufficient information for the audience in different viewing modes. The present application accurately identifies the viewing mode and interactive behavior of the audience, and adjusts the level and content of information display in real time, so that each audience can obtain the information most suitable for their needs. Experimental data shows that in the fast-browsing mode, the information display reduces 18.6% of the elements, successfully avoiding information overload; in the long-time stay mode, 32.5% of the interactive data display is increased, improving the audience's sense of participation in the content.

[0162] Secondly, the automatic screening and dynamic display of bullet screen priority effectively improve the interactive value of bullet screen. In traditional live platforms, the content of bullet screen may become chaotic due to large amount of information, and it is difficult for the audience to quickly obtain valuable bullet screen information. Through emotion analysis and interaction frequency calculation, the system can accurately screen the bullet screen with positive emotions, high interaction frequency and high relevance to live content, and preferentially display these bullet screens, thereby improving the interaction quality of the audience. Experimental results show that after screening, the display ratio of high priority bullet screen is 28.7%, which is 16.4% higher than that of 12.3% before screening.

[0163] Finally, based on the automatic adjustment of device and network optimization mode, not only ensures the smooth playing in low bandwidth environment, but also reduces resource waste. When the network condition is poor, the system automatically reduces the resolution of images and videos, avoids unnecessary data transmission, ensures the smoothness of audience experience, and optimizes the bandwidth utilization. Experimental data shows that in low bandwidth environment, the data flow of video resolution reduction is reduced by about 22.8%, ensuring stable playing under bandwidth limitation.

[0164] In summary, the innovative design of the application in audience behavior analysis, information display level dynamic adjustment, and bullet screen priority screening makes the live content display more accurate and personalized, significantly improves the user experience and interaction effect. Through the verification of experimental data, it can be clearly seen that the application technology has obvious innovation, practicality and advantage compared with the prior art, can effectively optimize the live content display and interaction process, and has high application value.

[0165] It should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, which should be covered by the scope of the claims of the present application.

Claims

1. A method for displaying live room information, characterized in that, The application comprises the following steps: Collecting audience data and analyzing audience viewing behavior in real time; According to the audience viewing behavior, automatically matching the viewing mode and displaying different information levels for different viewing modes; Analyzing the audience's bullet screen and automatically filtering the bullet screen priority; According to the adjusted information level and bullet screen priority, dynamically displaying live content; Displaying different information levels for different viewing modes includes dividing multiple audience data into multiple layers according to viewing mode categories, with each layer corresponding to a viewing mode category; According to the distribution proportion of each viewing mode in the overall audience data, determine the sampling proportion of each layer, and randomly sample from each layer as a training set, which is expressed by the formula: ； ； wherein, represents the mth layer of the sampling ratio, whose value is equal to the proportion of the mth layer audience data in the overall audience data, represents the number of samples in the mth layer audience data, represents the total number of samples in the overall audience data, represents the final training set represents the mth layer data set , the sample set randomly sampled according to the proportion, RS represents random sampling, and M represents the level of audience data; The similarity between the behavior features of the audience is calculated using the Euclidean distance, which is expressed by the formula: ; wherein, denotes a weighted Euclidean distance, denotes the behavior feature of the th viewer, denotes the value of the th behavior feature of the th viewer, denotes the behavior feature of the th viewer, denotes the value of the th behavior feature of the th viewer, denotes the weight of the th behavior feature, denotes the total dimension of the behavior features; Select , calculate the similarity between the behavior characteristics of each audience and other audiences, find the most similar audiences to the current audience, build an approximate KNN model, and the formula is represented as: ； wherein, represents the set of neighbors most similar to the viewer represents the set of neighbors most similar to the viewer represents selecting the neighbors most similar to the viewer represents selecting the neighbors most similar to the viewer represents the weighted Euclidean distance between the th viewer and the th viewer; represents selecting from the behavior features of all viewers in the training set represents selecting from the behavior features of all viewers in the training set​​ The construction of the approximate KNN model includes preprocessing the training set using the local sensitive hash data index structure, and training the approximate KNN model based on the training set after hierarchical sampling; For new viewers entering the live stream, their behavioral characteristics are extracted in real time, and then analyzed using an approximate KNN model. The viewing pattern categories of the nearest neighbors are used to determine the viewing pattern of new viewers through majority voting, as expressed by the formula: ; wherein, represents the predicted viewing pattern of a new viewer, MV represents the majority vote, represents the viewing pattern category of the th viewer; represents that the viewing pattern of the new viewer is selected from the most similar neighbors of the new viewer to vote. ​ 2. The live room information presentation method of claim 1, wherein: The audience data includes viewing behavior data, interactive behavior data, emotional and emotional data, and device and network data; The real-time analysis of audience viewing behavior includes extracting behavior characteristics from audience data; The behavior characteristics include user activity characteristics, viewing mode characteristics, emotional tendency characteristics, user type characteristics, participation characteristics, and device and network characteristics; The user activity characteristics include audience online duration and interaction frequency; The viewing mode characteristics include dwell time and browsing speed; The emotional tendency characteristics include emotional score and user type characteristics; The participation characteristics include interaction depth and interaction breadth; The device and network characteristics include device type and network status; Define the viewing mode based on the behavior characteristics; The viewing mode includes fast browsing mode, long-time staying mode, high interaction mode, low interaction mode, and device and network optimization mode.

3. The live room information presentation method of claim 2, wherein: Displaying different information levels for different viewing modes also includes automatically switching to the basic information level when the system detects that the audience is in the fast browsing mode, displaying the most core basic information, and reducing interface elements; When the audience stays on the live page for a long time, the system automatically switches to the intermediate information level and displays more interactive data and discussion topics; When the system detects that the audience is in the high interaction mode, the system switches to the deep information level and displays detailed interactive information and content details; When the system detects that the audience is in the low interaction mode, the system maintains the intermediate information level and provides interactive information to encourage the audience to increase interaction; When the system detects that the audience is in the device and network optimization mode, reduce the display of high-resolution images and videos in low-bandwidth environments.

4. The live room information presentation method of claim 3, wherein: Analyzing the audience's bullet screen includes sentiment analysis, interaction frequency calculation, and content relevance analysis; The sentiment analysis includes using natural language processing technology to analyze the sentiment of all audience bullet screens; The sentiment score of each bullet screen is divided into positive, negative and neutral categories; The interaction frequency calculation includes counting the interaction frequency of each bullet screen, and the bullet screen with high interaction frequency represents higher interaction value; The interaction frequency includes likes, forwards and replies; The content relevance analysis includes determining the relevance of each bullet screen to the current live content through semantic analysis technology.

5. The live room information presentation method of claim 4, wherein: The automatic screening of bullet screen priority includes designing a comprehensive scoring formula that takes into account sentiment analysis, interaction frequency, and content relevance, and the formula is expressed as: ； wherein, represents a comprehensive score, represents a decay factor, represents a time variable, represents a barrage at a time emotional score; represents a barrage at a time interaction frequency; represents a barrage relevance score with a current live content theme ; The range of values of the comprehensive score ;​ When , it indicates that the barrage has high interaction value, positive emotion, and high relevance to the live content, and is a high-priority barrage. When , it indicates that the barrage interaction value is low, the emotion is negative, and the relevance to the live content is low, which is a low-priority barrage.

6. The live room information presentation method of claim 5, wherein: According to the adjusted information level and bullet screen priority, dynamically displaying live content includes continuously monitoring the behavior changes of the audience and the comprehensive score of the bullet screen to ensure that the information display is synchronized with the current state of the audience; When the viewing mode of the audience changes, automatically switch to the corresponding information level; When the comprehensive score of the bullet screen changes, real-time adjust the display priority of the bullet screen.

7. A live room information display system for implementing the live room information display method according to any one of claims 1-6, characterized in that, It includes: The data collection module collects audience data in real time to analyze audience viewing behavior; The information level module automatically matches the viewing mode according to the audience viewing behavior and displays different information levels for different viewing modes; The bullet screen analysis module analyzes the audience's bullet screen and automatically screens the bullet screen priority; The continuous adjustment module dynamically displays live content according to the adjusted information level and bullet screen priority.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that: The processor executes the computer program to realize the steps of the live room information display method in any one of claims 1-6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the live room information display method in any one of claims 1-6.

Images