CN112667890A - Live broadcast interaction user behavior analysis method and device and computer equipment - Google Patents
Live broadcast interaction user behavior analysis method and device and computer equipment Download PDFInfo
- Publication number
- CN112667890A CN112667890A CN202011551128.9A CN202011551128A CN112667890A CN 112667890 A CN112667890 A CN 112667890A CN 202011551128 A CN202011551128 A CN 202011551128A CN 112667890 A CN112667890 A CN 112667890A
- Authority
- CN
- China
- Prior art keywords
- user behavior
- information
- user
- live broadcast
- target user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a live broadcast interaction user behavior analysis method and device and computer equipment. In the method, a preset user behavior analysis model is obtained by training a plurality of behavior information conditions to perform user behavior monitoring of a plurality of user interest types on a target user behavior monitoring strategy, so that the user behavior can be monitored through the plurality of behavior information conditions, the problem of incomplete monitoring caused by monitoring the user behavior by adopting a single behavior information condition is avoided, the problem of inaccurate obtained user behavior monitoring results is improved or solved, and the accuracy of the user behavior monitoring results is improved. Based on the user behavior monitoring method, the user behavior global analysis is carried out on the target user behavior information according to the user behavior monitoring result, the user behavior can be analyzed in multiple aspects, the completeness of the user behavior analysis is ensured, and the user behavior analysis can be highly matched with the actual user condition.
Description
Technical Field
The present disclosure relates to the field of live broadcast interaction and user behavior analysis technologies, and in particular, to a method and an apparatus for user behavior analysis of live broadcast interaction, and a computer device.
Background
With the development of information networks, the popularity of network live broadcasting is increasing more and more, and then more and more users of different age groups can experience the wonderful of live broadcasting, however, in order to accurately push live videos in which users of different age groups are interested, user behaviors need to be analyzed before the live videos are pushed.
At present, the technology for analyzing the user behavior is to analyze the user behavior only through browsing data of a live broadcast platform, so that the user behavior cannot be analyzed in many aspects, and further the integrity of the user behavior analysis cannot be ensured and the situation of the user behavior cannot be matched with the actual user.
Disclosure of Invention
In order to solve the technical problems in the related art, the present disclosure provides a live broadcast interactive user behavior analysis method, apparatus and computer device.
The invention provides a user behavior analysis method for live broadcast interaction, which comprises the following steps:
determining target user behavior information, wherein the target user behavior information is related to the behavior habit information of the monitored user behavior information;
generating a plurality of target user behavior monitoring strategies of the target user behavior information;
acquiring preset user behavior analysis models corresponding to the target user behavior monitoring strategies respectively; the preset user behavior analysis model is obtained by training a plurality of behavior information conditions;
based on the preset user behavior analysis models corresponding to the target user behavior monitoring strategies, carrying out user behavior monitoring on the corresponding target user behavior monitoring strategies in a plurality of user interest types to obtain user behavior monitoring results corresponding to the target user behavior monitoring strategies;
and performing user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the target user behavior monitoring strategies.
Optionally, the determining target user behavior information includes: monitoring behavior habit information of the user behavior information; when the behavior habit information is abnormal, taking the user behavior information corresponding to the abnormal behavior habit information as the target user behavior information; correspondingly, the behavior habit information corresponding to the target user behavior information comprises a plurality of behavior habit information, and then the target user behavior monitoring strategies for generating the target user behavior information are obtained;
wherein the plurality of target user behavior monitoring policies for generating the target user behavior information include: monitoring the plurality of behavior habit information respectively to obtain user behavior monitoring strategies corresponding to the plurality of behavior habit information respectively; and taking the user behavior monitoring strategies corresponding to the behavior habit information as a plurality of target user behavior monitoring strategies of the target user behavior information.
Optionally, the method further includes determining the preset user behavior analysis model, where determining the preset user behavior analysis model includes: acquiring a plurality of behavior information conditions; determining a matching relationship between the plurality of behavior information conditions; training the plurality of behavior information conditions based on the matching relationship to obtain the preset user behavior analysis model.
Optionally, the performing, based on the preset user behavior analysis model corresponding to each of the target user behavior monitoring policies, user behavior monitoring of multiple user interest types on the corresponding target user behavior monitoring policy to obtain user behavior monitoring results corresponding to each of the target user behavior monitoring policies includes:
based on the preset user behavior analysis models corresponding to the target user behavior monitoring strategies, monitoring user interest information of the corresponding target user behavior monitoring strategies to obtain interest information monitoring results corresponding to the target user behavior monitoring strategies;
and when the interest information monitoring results corresponding to the target user behavior monitoring strategies are not matched with preset interest information, determining the user behavior monitoring results corresponding to the target user behavior monitoring strategies as first user interest information.
Optionally, the method further comprises:
when the interest information monitoring results corresponding to the target user behavior monitoring strategies are matched with the preset interest information, acquiring live broadcast browsing information corresponding to the target user behavior monitoring strategies, wherein the live broadcast browsing information corresponding to the target user behavior monitoring strategies is browsing information matched with the corresponding behavior habit information;
when information browsing differences occur in the live browsing information corresponding to the target user behavior monitoring strategies, determining user behavior monitoring results corresponding to the target user behavior monitoring strategies as second user interest information;
and when the live broadcast browsing information corresponding to the target user behavior monitoring strategies does not have information browsing difference, determining the user behavior monitoring result corresponding to each target user behavior monitoring strategy as third user interest information.
Optionally, when the preset user behavior analysis model comprises an analysis model corresponding to a keyword of a user search engine,
the method for monitoring the user behaviors of the plurality of user interest types of the corresponding target user behavior monitoring strategies based on the preset user behavior analysis models corresponding to the plurality of target user behavior monitoring strategies respectively to obtain the user behavior monitoring results corresponding to the plurality of target user behavior monitoring strategies respectively comprises the following steps:
performing live broadcast access quantity analysis on user access live broadcast information corresponding to a corresponding target user behavior monitoring strategy based on an analysis model corresponding to the user search engine keyword to obtain a live broadcast access quantity analysis result;
when the analysis model corresponding to the user search engine keywords is used for analyzing the live broadcast access time period of the analysis result of the live broadcast access amount and determining that the corresponding target user behavior monitoring strategy has access behaviors in the live broadcast access time period, the analysis model corresponding to the user search engine keywords is used for analyzing the live broadcast stay time period of the corresponding target user behavior monitoring strategy to obtain the analysis result of the live broadcast stay time period;
analyzing the analysis result of the live broadcast stay time period based on the analysis model corresponding to the user search engine keyword to obtain a user behavior monitoring result of a target user behavior monitoring strategy corresponding to the analysis model corresponding to the user search engine keyword;
wherein the performing, based on the user behavior monitoring results corresponding to the plurality of target user behavior monitoring policies, a user behavior global analysis on the target user behavior information includes:
performing global analysis on corresponding user behavior monitoring results based on preset user behavior analysis models corresponding to the target user behavior monitoring strategies to obtain global analysis results of user behaviors corresponding to the target user behavior monitoring strategies;
and sending the global analysis result of the corresponding user behavior to the information receiving terminal corresponding to the corresponding preset user behavior analysis model based on the preset user behavior analysis model corresponding to each target user behavior monitoring strategy.
Optionally, the method further comprises:
generating live broadcast user push content corresponding to the global analysis result according to the global analysis result of each target user behavior monitoring strategy;
the method for generating the live broadcast user push content corresponding to the global analysis result according to the global analysis result of each target user behavior monitoring strategy specifically comprises the following steps:
according to the global analysis result of each target user behavior monitoring strategy, acquiring preference labels of a plurality of user group preference information corresponding to a user group to be analyzed, wherein the plurality of user group preference information comprises the user group preference information corresponding to each user group in the user group to be analyzed;
determining a live user click preference based on preference tags of the plurality of user group preference information;
acquiring live click records of each user group in the user groups to be analyzed within a first preset time range;
determining click content corresponding to the live click record of each user group in a first preset time range;
based on a preset user click analysis model, carrying out live broadcast watching habit information analysis on the user groups in the user group to be analyzed according to the click content and the live broadcast user click preference to obtain live broadcast watching habit information among the user groups in the user group to be analyzed;
determining live broadcast interaction influence information among the user groups in the user group to be analyzed according to the live broadcast viewing habit information among the user groups in the user group to be analyzed;
generating live broadcast user push content corresponding to the global analysis result based on the live broadcast interaction influence information;
the preset user click analysis model comprises a live click recording marking unit, a live content counting unit, a live time period splitting unit and a live barrage identification unit;
the method comprises the following steps of based on a preset user click analysis model, carrying out live broadcast watching habit information analysis on a user group in a user group to be analyzed according to click content and live broadcast user click preference, and obtaining live broadcast watching habit information among the user group in the user group to be analyzed, wherein the live broadcast watching habit information comprises the following steps:
performing live click recording marking on the click content based on the live click recording marking unit to obtain a live click recording marking track;
performing live broadcast content statistics on the live broadcast click record marking track and the live broadcast user click preference based on the live broadcast content statistics unit to obtain a live broadcast content statistics result;
performing live broadcast content correction processing on the live broadcast content statistical result based on the live broadcast time interval splitting unit to obtain a processed live broadcast content statistical result;
determining live broadcast watching habit information on the basis of the live broadcast bullet screen recognition unit on the processed live broadcast content statistical result to obtain live broadcast watching habit information among user groups in the user group to be analyzed;
wherein the determining of the click preference of the live broadcast user based on the preference tags of the plurality of user group preference information comprises:
determining user group preference trend information in a second preset time range according to the preference labels of the plurality of user group preference information;
extracting the attention of the live broadcast content according to the preference trend information of the user group to obtain the attention level of the live broadcast content;
ordering the attention degree of the live broadcast content to obtain the click preference of the live broadcast user;
the determining of the preference trend information of the user groups in a second preset time range according to the preference labels of the plurality of user group preference information comprises the following steps:
determining preference trend identification information of the plurality of user group preference information according to the preference labels of the plurality of user group preference information;
constructing user group preference trend information within the second preset time range according to preference trend identification information of the plurality of user group preference information;
wherein the method further comprises: acquiring sample live broadcast user click preferences corresponding to a plurality of sample user group sets marked with live broadcast watching habit information among user groups and corresponding sample click contents;
based on the click preferences of the sample live users corresponding to the sample user group sets marked with the live viewing habit information among the user groups and the corresponding sample click contents, training of analyzing the live viewing habit information on a preset neural network model, and adjusting model parameters of the preset neural network model in the training of analyzing the live viewing habit information until the preset neural network model meets a preset convergence condition to obtain a preset user click analysis model;
the obtaining of the sample click content corresponding to the sample user group set marked with the live viewing habit information among the user groups includes:
respectively obtaining a sample live broadcast click record of each user group in the sample user group set within a third preset time range;
respectively carrying out live broadcast duration analysis on the sample live broadcast click records to obtain corresponding sample click contents;
wherein, the determining the live broadcast interaction influence information among the user groups in the user group to be analyzed according to the live broadcast watching habit information among the user groups in the user group to be analyzed comprises:
and when the live broadcast watching habit information is matched with a preset watching condition, determining live broadcast interaction influence information among user groups in the user group to be analyzed.
The invention also provides a user behavior analysis device for live broadcast interaction, which comprises:
the behavior information determining module is used for determining behavior information of a target user, wherein the behavior information of the target user is related to the behavior habit information of the monitored user behavior information;
the monitoring strategy generating module is used for generating a plurality of target user behavior monitoring strategies of the target user behavior information;
the analysis model acquisition module is used for acquiring preset user behavior analysis models corresponding to the target user behavior monitoring strategies; the preset user behavior analysis model is obtained by training a plurality of behavior information conditions;
the monitoring strategy monitoring module is used for carrying out user behavior monitoring of a plurality of user interest types on the corresponding target user behavior monitoring strategies based on the preset user behavior analysis models corresponding to the target user behavior monitoring strategies respectively to obtain user behavior monitoring results corresponding to the target user behavior monitoring strategies respectively;
and the user behavior analysis module is used for carrying out user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the target user behavior monitoring strategies.
The invention also provides a computer device comprising a processor and a memory communicating with each other, the processor being configured to retrieve a computer program from the memory and to implement the method of any one of the preceding claims by running the computer program.
The invention also provides a computer-readable storage medium having stored thereon a computer program which, when run, implements the method of any of the above.
The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects.
The invention provides a live broadcast interactive user behavior analysis method, a live broadcast interactive user behavior analysis device and computer equipment, wherein a plurality of behavior information conditions are trained in advance to obtain a preset user behavior analysis model, when user behavior global analysis is carried out on target user behavior information, a plurality of target user behavior monitoring strategies are generated according to the monitored target user behavior information, and then the target user behavior monitoring strategies are distributed to the corresponding preset user behavior analysis model for processing, so that user behavior is monitored according to a plurality of user interest types to obtain user behavior monitoring results, and on the basis, the target user behavior information is subjected to user behavior global analysis based on a plurality of user monitoring results.
Therefore, the user behavior monitoring of multiple user interest types is carried out on the target user behavior monitoring strategy by the preset user behavior analysis model obtained through training of the multiple behavior information conditions, so that the user behavior can be monitored through the multiple behavior information conditions, the problem that monitoring is incomplete due to the fact that the user behavior is monitored through a single behavior information condition is avoided, the problem that the obtained user behavior monitoring result is inaccurate is further improved or solved, and the accuracy of the user behavior monitoring result is improved. Based on the user behavior monitoring method, the user behavior global analysis is carried out on the target user behavior information according to the user behavior monitoring result, the user behavior can be analyzed in multiple aspects, the completeness of the user behavior analysis is ensured, and the user behavior analysis can be highly matched with the actual user condition.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 is a flowchart of a user behavior analysis method for live broadcast interaction according to an embodiment of the present invention.
Fig. 2 is a block diagram of a live interaction user behavior analysis apparatus according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of a hardware structure of a computer device according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.
Referring to fig. 1, a flow chart of a live broadcast interactive user behavior analysis method is provided, and when the method is implemented, the following contents described in steps S11 to S15 are specifically executed.
And step S11, determining target user behavior information.
In this embodiment, the target user behavior information is related to the behavior habit information of the monitored user behavior information.
Step S12, generating a plurality of target user behavior monitoring policies of the target user behavior information.
In this embodiment, the target user behavior monitoring policy represents a manner of monitoring the target user behavior or monitoring instructional information.
Step S13, obtaining preset user behavior analysis models corresponding to the target user behavior monitoring policies.
In this embodiment, the preset user behavior analysis model is obtained by training a plurality of behavior information conditions. The behavior information condition is a condition set by performing behavior characteristic limitation on the user behavior information.
Step S14, based on the preset user behavior analysis model corresponding to each of the target user behavior monitoring policies, performing user behavior monitoring of multiple user interest types on the corresponding target user behavior monitoring policy, to obtain user behavior monitoring results corresponding to each of the target user behavior monitoring policies.
Step S15, performing user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the multiple target user behavior monitoring policies.
The method described in the steps S11-S15 can achieve the following beneficial effects: the method comprises the steps of training by a plurality of behavior information conditions in advance to obtain a preset user behavior analysis model, generating a plurality of target user behavior monitoring strategies according to monitored target user behavior information when performing user behavior global analysis on the target user behavior information, distributing the target user behavior monitoring strategies to the corresponding preset user behavior analysis model for processing, monitoring user behaviors according to a plurality of user interest types to obtain user behavior monitoring results, and performing user behavior global analysis on the target user behavior information on the basis of the user behavior monitoring results.
Therefore, the user behavior monitoring of multiple user interest types is carried out on the target user behavior monitoring strategy by the preset user behavior analysis model obtained through training of the multiple behavior information conditions, so that the user behavior can be monitored through the multiple behavior information conditions, the problem that monitoring is incomplete due to the fact that the user behavior is monitored through a single behavior information condition is avoided, the problem that the obtained user behavior monitoring result is inaccurate is further improved or solved, and the accuracy of the user behavior monitoring result is improved. Based on the user behavior monitoring method, the user behavior global analysis is carried out on the target user behavior information according to the user behavior monitoring result, the user behavior can be analyzed in multiple aspects, the completeness of the user behavior analysis is ensured, and the user behavior analysis can be highly matched with the actual user condition.
In a specific implementation, the determining target user behavior information described in step S11 includes:
monitoring behavior habit information of the user behavior information; and when the behavior habit information is abnormal, taking the user behavior information corresponding to the abnormal behavior habit information as the target user behavior information.
Further, in this embodiment, the behavior habit information corresponding to the target user behavior information includes a plurality of behavior habit information, and the target user behavior monitoring policy that generates the target user behavior information is a plurality of target user behavior monitoring policies.
On the basis of step S11, the multiple target user behavior monitoring policies for generating the target user behavior information described in step S12 may specifically include step S121 and step S122.
Step S121, monitoring the behavior habit information respectively, to obtain a user behavior monitoring policy corresponding to each of the behavior habit information.
Step S122, using the user behavior monitoring policy corresponding to each of the behavior habit information as a plurality of target user behavior monitoring policies of the target user behavior information.
When the content is executed, the user behavior information corresponding to the abnormal behavior habit information is used as the target user behavior information, so that the follow-up analysis on the user behavior can be more effective, and meanwhile, the one-to-one correspondence between each piece of behavior habit information and the user behavior monitoring strategy can be ensured.
Further, the method further comprises: and determining the preset user behavior analysis model.
It can be understood that the determining of the preset user behavior analysis model specifically includes steps a to C:
and step A, acquiring a plurality of behavior information conditions.
And B, determining the matching relation among the plurality of behavior information conditions.
And step C, training the plurality of behavior information conditions based on the matching relation to obtain the preset user behavior analysis model.
According to the contents described in the steps A to C, a preset user behavior analysis model can be quickly trained according to a plurality of behavior information conditions, so that the target user behavior monitoring strategy can be comprehensively analyzed through the plurality of behavior information conditions in the preset user behavior analysis model.
In an alternative embodiment, in order to improve the accuracy of obtaining the user behavior monitoring result, the preset user behavior analysis model described in step S14 is used to perform user behavior monitoring of multiple user interest types on the corresponding target user behavior monitoring policy based on the corresponding target user behavior monitoring policy, so as to obtain the user behavior monitoring result corresponding to each target user behavior monitoring policy, which may specifically include the contents described in step S141 and step S142.
Step S141, based on the preset user behavior analysis model corresponding to each of the target user behavior monitoring policies, performing user interest information monitoring on the corresponding target user behavior monitoring policy to obtain interest information monitoring results corresponding to each of the target user behavior monitoring policies.
Step S142, when the interest information monitoring results corresponding to the target user behavior monitoring policies are not matched with preset interest information, determining that the user behavior monitoring results corresponding to the target user behavior monitoring policies are first user interest information.
Executing the contents described in step S141 and step S142, performing user behavior monitoring of multiple user interest types in one-to-one correspondence with multiple target user behavior monitoring policies, so as to ensure that the obtained user behavior monitoring results are relatively comprehensive, further matching the obtained interest information monitoring results with preset interest information, and further determining the user behavior monitoring results corresponding to the multiple target user behavior monitoring policies under the condition that the interest information monitoring results are not matched with the preset interest information, so as to determine the user behavior monitoring results corresponding to the target user behavior monitoring policies under the targeted condition, thereby improving the accuracy of obtaining the user behavior monitoring results.
It can be understood that when the interest information monitoring results corresponding to the target user behavior monitoring policies are matched with the preset interest information, live browsing information corresponding to the target user behavior monitoring policies is obtained, and the live browsing information corresponding to the target user behavior monitoring policies is browsing information matched with the corresponding behavior habit information;
when information browsing differences occur in the live browsing information corresponding to the target user behavior monitoring strategies, determining user behavior monitoring results corresponding to the target user behavior monitoring strategies as second user interest information;
and when the live broadcast browsing information corresponding to the target user behavior monitoring strategies does not have information browsing difference, determining the user behavior monitoring result corresponding to each target user behavior monitoring strategy as third user interest information.
In an alternative embodiment, when the preset user behavior analysis model includes an analysis model corresponding to a user search engine keyword. Further, the step S14 may include performing user behavior monitoring on a plurality of user interest types for the corresponding target user behavior monitoring policy based on the preset user behavior analysis model corresponding to each of the target user behavior monitoring policies to obtain user behavior monitoring results corresponding to each of the target user behavior monitoring policies, and may further include the contents described in the steps S143 to S145.
And S143, performing live broadcast access quantity analysis on the user access live broadcast information corresponding to the corresponding target user behavior monitoring strategy based on the analysis model corresponding to the user search engine keyword to obtain a live broadcast access quantity analysis result.
Step S144, when the analysis model corresponding to the user search engine keywords is used for analyzing the live broadcast access time period of the analysis result of the live broadcast access amount and determining that the corresponding target user behavior monitoring strategy has access behaviors in the live broadcast access time period, the analysis model corresponding to the user search engine keywords is used for analyzing the live broadcast stay time period of the corresponding target user behavior monitoring strategy to obtain the analysis result of the live broadcast stay time period.
Step S145, analyzing the analysis result of the live broadcast stay time period based on the analysis model corresponding to the user search engine keyword to obtain the user behavior monitoring result of the target user behavior monitoring strategy corresponding to the analysis model corresponding to the user search engine keyword;
and executing the contents described in the steps S143 to S145, and performing live broadcast access amount analysis on the live broadcast information accessed by the user according to the user search engine keywords, so that on one hand, the live broadcast access amount can be analyzed, and on the other hand, the live broadcast heat can be analyzed. And then carry out the analysis of the time slot of the live broadcast visit according to the keyword of the search engine of the user on the analysis result of the live broadcast visit, thus can guarantee the time slot that the volume of the live broadcast visit is higher or lower, on this basis, carry out the analysis of the analysis result of the live broadcast stay time slot according to the keyword of the search engine of the user, so, can confirm the stay time slot when the user watches the live broadcast, and then can judge the love degree of the user to the live broadcast, therefore, can carry on the comprehensive monitoring to the user behavior through the above-mentioned monitoring mode.
In an alternative embodiment, in order to perform multi-aspect analysis on the user behavior, ensure the integrity of the user behavior analysis, and ensure that the user behavior analysis can be highly matched with the actual user situation, and avoid losing the global analysis result, the step S15 may perform the user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the plurality of target user behavior monitoring policies, which specifically includes the contents described in step S151 and step S152.
Step S151, performing global analysis on corresponding user behavior monitoring results based on the preset user behavior analysis models corresponding to the target user behavior monitoring policies, to obtain global analysis results of user behaviors corresponding to the target user behavior monitoring policies.
Step S152, based on the preset user behavior analysis models corresponding to the target user behavior monitoring policies, sending the global analysis result of the corresponding user behavior to the information receiving terminal corresponding to the corresponding preset user behavior analysis model. In this embodiment, the information receiving terminal may be a radio computer, a mobile phone, a tablet, or the like.
The contents described in step S151 and step S152 are executed, the user behavior monitoring result is globally analyzed, the user behavior can be analyzed in many ways, the integrity of the user behavior analysis is ensured, and the user behavior analysis can be highly matched with the actual user condition. And then send the overall situation analysis result to the information receiving terminal in order to keep, can avoid overall situation analysis result to lose like this, facilitate the follow-up to look for at the same time.
In an alternative embodiment, the method further comprises step S16: and generating live broadcast user push content corresponding to the global analysis result according to the global analysis result of each target user behavior monitoring strategy.
In a specific implementation, in order to push live content according to the viewing preference and the viewing requirement of the user and further save the time for the user to search for the live content, the live user push content corresponding to the global analysis result is generated according to the global analysis result of each target user behavior monitoring policy described in step S16, and may specifically include the content described in steps S161 to S167.
Step S161, obtaining preference labels of multiple user group preference information corresponding to the user group to be analyzed according to the global analysis result of each target user behavior monitoring policy, where the multiple user group preference information includes user group preference information corresponding to each user group in the user group to be analyzed.
And step S162, determining the click preference of the live broadcast user based on the preference labels of the preference information of the plurality of user groups.
Step S163, acquiring a live click record of each user group in the user groups to be analyzed within a first preset time range.
Step S164, determining click content corresponding to the live click record of each user group in a first preset time range.
And S165, based on a preset user click analysis model, carrying out live broadcast watching habit information analysis on the user group in the user group to be analyzed according to the click content and the live broadcast user click preference to obtain the live broadcast watching habit information among the user groups in the user group to be analyzed.
Step S166, determining the live broadcast interaction influence information among the user groups in the user group to be analyzed according to the live broadcast watching habit information among the user groups in the user group to be analyzed.
And step S167, generating live broadcast user push content corresponding to the global analysis result based on the live broadcast interaction influence information.
Therefore, the contents described in the steps S161 to S167 are executed, according to the click preference and the click content of the live broadcast user, the live broadcast watching habit information analysis is performed on the user group in the user group to be analyzed, the live broadcast watching habit information among the user group is determined, and further the live broadcast interaction influence information among the user group in the user group to be analyzed is determined according to the live broadcast watching habit information.
In an alternative embodiment, the preset user click analysis model includes a live click record marking unit, a live content statistics unit, a live time period splitting unit, and a live barrage recognition unit. In this embodiment, the user click analysis model may be a network model that is trained in advance, and the corresponding units may be different functional layers of the network model.
It can be understood that, in order to quickly analyze the live viewing habit information among the user groups, the live viewing habit information analysis is performed on the user groups in the user group to be analyzed based on the preset user click analysis model described in step S165 according to the click content and the live user click preference, so as to obtain the live viewing habit information among the user groups in the user group to be analyzed, where the live viewing habit information specifically includes the content described in step S1651 to step S1654.
Step S1651, based on the live click record marking unit, live click record marking is carried out on the click content, and a live click record marking track is obtained.
Step S1652, based on the live broadcast content statistical unit, live broadcast content statistics is carried out on the live broadcast click record mark track and the live broadcast user click preference, and a live broadcast content statistical result is obtained.
Step S1653, based on the live broadcast time interval splitting unit, live broadcast content correction processing is carried out on the live broadcast content statistical result, and a processed live broadcast content statistical result is obtained.
Step S1654, based on the live broadcast barrage identification unit, live broadcast watching habit information determination is carried out on the processed live broadcast content statistical result, and live broadcast watching habit information among user groups in the user group to be analyzed is obtained.
And executing the contents described in the steps S1651 to S1654, firstly, the live click recording marking unit carries out live click recording marking on the click contents to determine a live click recording marking track, and then live content statistics is carried out on the live click recording marking track live user click preference according to the live content statistics unit to count a live content statistics result, so that live content correction is carried out on the live content statistics result according to time intervals to ensure the accuracy of the live content statistics result. Omission is avoided in statistics, on the basis, live broadcast watching habit information is determined according to a live broadcast content statistical result of the live broadcast bullet screen recognition unit after processing, and therefore the live broadcast watching habit information among user groups can be quickly analyzed.
Further, the preference tag described in step S162 based on the plurality of user group preference information determines a live user click preference, which specifically includes the contents described in step S1621 to step S1623.
Step S1621, determining user group preference trend information within a second preset time range according to the preference tags of the plurality of user group preference information.
And S1622, extracting the attention of the live broadcast content according to the preference trend information of the user group to obtain the attention level of the live broadcast content.
And step S1623, performing attention ranking on the attention levels of the live broadcast contents to obtain click preferences of the live broadcast users.
Therefore, the attention of the live broadcast content can be extracted according to the preference trend information of the user group, and the attention of the user to the live broadcast content is further extracted, so that the attention levels of the live broadcast content are sorted, and the click preference of the live broadcast user can be visually and clearly checked.
It is to be understood that the determining of the preference tendency information of the user group in the second preset time range according to the preference labels of the plurality of user group preference information described in step S1621 includes step S16211 and step S16212.
Step S16211, determining preference trend identification information of the plurality of user group preference information according to the preference labels of the plurality of user group preference information.
Step S16212, constructing the user group preference trend information within the second preset time range according to the preference trend identification information of the plurality of user group preference information.
In an alternative embodiment, the method further comprises: acquiring sample live broadcast user click preferences corresponding to a plurality of sample user group sets marked with live broadcast watching habit information among user groups and corresponding sample click contents;
based on the click preferences of the sample live users corresponding to the sample user group sets marked with the live viewing habit information among the user groups and the corresponding sample click contents, training of analyzing the live viewing habit information on a preset neural network model, and adjusting model parameters of the preset neural network model in the training of analyzing the live viewing habit information until the preset neural network model meets a preset convergence condition to obtain a preset user click analysis model;
the obtaining of the sample click content corresponding to the sample user group set marked with the live viewing habit information among the user groups includes:
respectively obtaining a sample live broadcast click record of each user group in the sample user group set within a third preset time range;
respectively carrying out live broadcast duration analysis on the sample live broadcast click records to obtain corresponding sample click contents;
further, the determining, according to the live viewing habit information among the user groups in the user group to be analyzed, live interaction influence information among the user groups in the user group to be analyzed in step S166 specifically includes: and when the live broadcast watching habit information is matched with a preset watching condition, determining live broadcast interaction influence information among user groups in the user group to be analyzed.
Based on the same inventive concept, please refer to fig. 2, the present invention further provides a block diagram of a live interaction user behavior analysis apparatus 20, which may include the following functional modules.
The behavior information determining module 21 is configured to determine target user behavior information, where the target user behavior information is related to the behavior habit information of the monitored user behavior information.
And the monitoring policy generating module 22 is configured to generate a plurality of target user behavior monitoring policies of the target user behavior information.
The analysis model obtaining module 23 is configured to obtain preset user behavior analysis models corresponding to the target user behavior monitoring policies; the preset user behavior analysis model is obtained by training a plurality of behavior information conditions.
The monitoring policy monitoring module 24 is configured to perform user behavior monitoring of multiple user interest types on the corresponding target user behavior monitoring policies based on preset user behavior analysis models corresponding to the multiple target user behavior monitoring policies, so as to obtain user behavior monitoring results corresponding to the multiple target user behavior monitoring policies.
And the user behavior analysis module 25 is configured to perform user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the multiple target user behavior monitoring policies.
On the basis of the above, please refer to fig. 3 in combination, there is provided a computer device 110, which includes a processor 111, and a memory 112 and a bus 113 connected to the processor 111; wherein, the processor 111 and the memory 112 complete the communication with each other through the bus 113; the processor 111 is used to call program instructions in the memory 112 to perform the above-described method.
Further, a readable storage medium is provided, on which a program is stored, which when executed by a processor implements the method described above.
It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.
Claims (10)
1. A method for analyzing user behavior of live broadcast interaction is characterized by comprising the following steps:
determining target user behavior information, wherein the target user behavior information is related to the behavior habit information of the monitored user behavior information;
generating a plurality of target user behavior monitoring strategies of the target user behavior information;
acquiring preset user behavior analysis models corresponding to the target user behavior monitoring strategies respectively; the preset user behavior analysis model is obtained by training a plurality of behavior information conditions;
based on the preset user behavior analysis models corresponding to the target user behavior monitoring strategies, carrying out user behavior monitoring on the corresponding target user behavior monitoring strategies in a plurality of user interest types to obtain user behavior monitoring results corresponding to the target user behavior monitoring strategies;
and performing user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the target user behavior monitoring strategies.
2. The method of claim 1, wherein the determining target user behavior information comprises: monitoring behavior habit information of the user behavior information; when the behavior habit information is abnormal, taking the user behavior information corresponding to the abnormal behavior habit information as the target user behavior information; correspondingly, the behavior habit information corresponding to the target user behavior information comprises a plurality of behavior habit information, and then the target user behavior monitoring strategies for generating the target user behavior information are obtained;
wherein the plurality of target user behavior monitoring policies for generating the target user behavior information include: monitoring the plurality of behavior habit information respectively to obtain user behavior monitoring strategies corresponding to the plurality of behavior habit information respectively; and taking the user behavior monitoring strategies corresponding to the behavior habit information as a plurality of target user behavior monitoring strategies of the target user behavior information.
3. The method according to claim 1 or 2, further comprising determining the preset user behavior analysis model, the determining the preset user behavior analysis model comprising: acquiring a plurality of behavior information conditions; determining a matching relationship between the plurality of behavior information conditions; training the plurality of behavior information conditions based on the matching relationship to obtain the preset user behavior analysis model.
4. The method according to claim 2, wherein the step of performing user behavior monitoring of multiple user interest types on the corresponding target user behavior monitoring policy based on a preset user behavior analysis model corresponding to each of the multiple target user behavior monitoring policies to obtain user behavior monitoring results corresponding to each of the multiple target user behavior monitoring policies comprises:
based on the preset user behavior analysis models corresponding to the target user behavior monitoring strategies, monitoring user interest information of the corresponding target user behavior monitoring strategies to obtain interest information monitoring results corresponding to the target user behavior monitoring strategies;
and when the interest information monitoring results corresponding to the target user behavior monitoring strategies are not matched with preset interest information, determining the user behavior monitoring results corresponding to the target user behavior monitoring strategies as first user interest information.
5. The method of claim 4, further comprising:
when the interest information monitoring results corresponding to the target user behavior monitoring strategies are matched with the preset interest information, acquiring live broadcast browsing information corresponding to the target user behavior monitoring strategies, wherein the live broadcast browsing information corresponding to the target user behavior monitoring strategies is browsing information matched with the corresponding behavior habit information;
when information browsing differences occur in the live browsing information corresponding to the target user behavior monitoring strategies, determining user behavior monitoring results corresponding to the target user behavior monitoring strategies as second user interest information;
and when the live broadcast browsing information corresponding to the target user behavior monitoring strategies does not have information browsing difference, determining the user behavior monitoring result corresponding to each target user behavior monitoring strategy as third user interest information.
6. The method of claim 1, wherein when the predetermined analysis model of user behavior comprises an analysis model corresponding to a keyword of a user search engine,
the method for monitoring the user behaviors of the plurality of user interest types of the corresponding target user behavior monitoring strategies based on the preset user behavior analysis models corresponding to the plurality of target user behavior monitoring strategies respectively to obtain the user behavior monitoring results corresponding to the plurality of target user behavior monitoring strategies respectively comprises the following steps:
performing live broadcast access quantity analysis on user access live broadcast information corresponding to a corresponding target user behavior monitoring strategy based on an analysis model corresponding to the user search engine keyword to obtain a live broadcast access quantity analysis result;
when the analysis model corresponding to the user search engine keywords is used for analyzing the live broadcast access time period of the analysis result of the live broadcast access amount and determining that the corresponding target user behavior monitoring strategy has access behaviors in the live broadcast access time period, the analysis model corresponding to the user search engine keywords is used for analyzing the live broadcast stay time period of the corresponding target user behavior monitoring strategy to obtain the analysis result of the live broadcast stay time period;
analyzing the analysis result of the live broadcast stay time period based on the analysis model corresponding to the user search engine keyword to obtain a user behavior monitoring result of a target user behavior monitoring strategy corresponding to the analysis model corresponding to the user search engine keyword;
wherein the performing, based on the user behavior monitoring results corresponding to the plurality of target user behavior monitoring policies, a user behavior global analysis on the target user behavior information includes:
performing global analysis on corresponding user behavior monitoring results based on preset user behavior analysis models corresponding to the target user behavior monitoring strategies to obtain global analysis results of user behaviors corresponding to the target user behavior monitoring strategies;
and sending the global analysis result of the corresponding user behavior to the information receiving terminal corresponding to the corresponding preset user behavior analysis model based on the preset user behavior analysis model corresponding to each target user behavior monitoring strategy.
7. The method of claim 6, further comprising:
generating live broadcast user push content corresponding to the global analysis result according to the global analysis result of each target user behavior monitoring strategy;
the method for generating the live broadcast user push content corresponding to the global analysis result according to the global analysis result of each target user behavior monitoring strategy specifically comprises the following steps:
according to the global analysis result of each target user behavior monitoring strategy, acquiring preference labels of a plurality of user group preference information corresponding to a user group to be analyzed, wherein the plurality of user group preference information comprises the user group preference information corresponding to each user group in the user group to be analyzed;
determining a live user click preference based on preference tags of the plurality of user group preference information;
acquiring live click records of each user group in the user groups to be analyzed within a first preset time range;
determining click content corresponding to the live click record of each user group in a first preset time range;
based on a preset user click analysis model, carrying out live broadcast watching habit information analysis on the user groups in the user group to be analyzed according to the click content and the live broadcast user click preference to obtain live broadcast watching habit information among the user groups in the user group to be analyzed;
determining live broadcast interaction influence information among the user groups in the user group to be analyzed according to the live broadcast viewing habit information among the user groups in the user group to be analyzed;
generating live broadcast user push content corresponding to the global analysis result based on the live broadcast interaction influence information;
the preset user click analysis model comprises a live click recording marking unit, a live content counting unit, a live time period splitting unit and a live barrage identification unit;
the method comprises the following steps of based on a preset user click analysis model, carrying out live broadcast watching habit information analysis on a user group in a user group to be analyzed according to click content and live broadcast user click preference, and obtaining live broadcast watching habit information among the user group in the user group to be analyzed, wherein the live broadcast watching habit information comprises the following steps:
performing live click recording marking on the click content based on the live click recording marking unit to obtain a live click recording marking track;
performing live broadcast content statistics on the live broadcast click record marking track and the live broadcast user click preference based on the live broadcast content statistics unit to obtain a live broadcast content statistics result;
performing live broadcast content correction processing on the live broadcast content statistical result based on the live broadcast time interval splitting unit to obtain a processed live broadcast content statistical result;
determining live broadcast watching habit information on the basis of the live broadcast bullet screen recognition unit on the processed live broadcast content statistical result to obtain live broadcast watching habit information among user groups in the user group to be analyzed;
wherein the determining of the click preference of the live broadcast user based on the preference tags of the plurality of user group preference information comprises:
determining user group preference trend information in a second preset time range according to the preference labels of the plurality of user group preference information;
extracting the attention of the live broadcast content according to the preference trend information of the user group to obtain the attention level of the live broadcast content;
ordering the attention degree of the live broadcast content to obtain the click preference of the live broadcast user;
the determining of the preference trend information of the user groups in a second preset time range according to the preference labels of the plurality of user group preference information comprises the following steps:
determining preference trend identification information of the plurality of user group preference information according to the preference labels of the plurality of user group preference information;
constructing user group preference trend information within the second preset time range according to preference trend identification information of the plurality of user group preference information;
wherein the method further comprises: acquiring sample live broadcast user click preferences corresponding to a plurality of sample user group sets marked with live broadcast watching habit information among user groups and corresponding sample click contents;
based on the click preferences of the sample live users corresponding to the sample user group sets marked with the live viewing habit information among the user groups and the corresponding sample click contents, training of analyzing the live viewing habit information on a preset neural network model, and adjusting model parameters of the preset neural network model in the training of analyzing the live viewing habit information until the preset neural network model meets a preset convergence condition to obtain a preset user click analysis model;
the obtaining of the sample click content corresponding to the sample user group set marked with the live viewing habit information among the user groups includes:
respectively obtaining a sample live broadcast click record of each user group in the sample user group set within a third preset time range;
respectively carrying out live broadcast duration analysis on the sample live broadcast click records to obtain corresponding sample click contents;
wherein, the determining the live broadcast interaction influence information among the user groups in the user group to be analyzed according to the live broadcast watching habit information among the user groups in the user group to be analyzed comprises:
and when the live broadcast watching habit information is matched with a preset watching condition, determining live broadcast interaction influence information among user groups in the user group to be analyzed.
8. An apparatus for analyzing user behavior of live interaction, the apparatus comprising:
the behavior information determining module is used for determining behavior information of a target user, wherein the behavior information of the target user is related to the behavior habit information of the monitored user behavior information;
the monitoring strategy generating module is used for generating a plurality of target user behavior monitoring strategies of the target user behavior information;
the analysis model acquisition module is used for acquiring preset user behavior analysis models corresponding to the target user behavior monitoring strategies; the preset user behavior analysis model is obtained by training a plurality of behavior information conditions;
the monitoring strategy monitoring module is used for carrying out user behavior monitoring of a plurality of user interest types on the corresponding target user behavior monitoring strategies based on the preset user behavior analysis models corresponding to the target user behavior monitoring strategies respectively to obtain user behavior monitoring results corresponding to the target user behavior monitoring strategies respectively;
and the user behavior analysis module is used for carrying out user behavior global analysis on the target user behavior information based on the user behavior monitoring results corresponding to the target user behavior monitoring strategies.
9. A computer device comprising a processor and a memory in communication with each other, the processor being configured to retrieve a computer program from the memory and to implement the method of any one of claims 1 to 7 by running the computer program.
10. A computer-readable storage medium, on which a computer program is stored which, when executed, implements the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011551128.9A CN112667890A (en) | 2020-12-24 | 2020-12-24 | Live broadcast interaction user behavior analysis method and device and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011551128.9A CN112667890A (en) | 2020-12-24 | 2020-12-24 | Live broadcast interaction user behavior analysis method and device and computer equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112667890A true CN112667890A (en) | 2021-04-16 |
Family
ID=75409999
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011551128.9A Withdrawn CN112667890A (en) | 2020-12-24 | 2020-12-24 | Live broadcast interaction user behavior analysis method and device and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112667890A (en) |
-
2020
- 2020-12-24 CN CN202011551128.9A patent/CN112667890A/en not_active Withdrawn
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9462313B1 (en) | Prediction of media selection consumption using analysis of user behavior | |
| US11914639B2 (en) | Multimedia resource matching method and apparatus, storage medium, and electronic apparatus | |
| CN112364202B (en) | Video recommendation method and device and electronic equipment | |
| CN113283185B (en) | Federal model training and client imaging method, device, equipment and medium | |
| CN110543598A (en) | information recommendation method and device and terminal | |
| CN110163376B (en) | Sample detection method, media object identification method, device, terminal and medium | |
| CN116797282B (en) | Real-time monitoring system and monitoring method for advertisement delivery | |
| CN113407773A (en) | Short video intelligent recommendation method and system, electronic device and storage medium | |
| CN110941738A (en) | Recommendation method and device, electronic equipment and computer-readable storage medium | |
| CN108376158B (en) | Student mobile terminal behavior log analysis method, device and system | |
| CN111177725B (en) | Method, device, equipment and storage medium for detecting malicious click operation | |
| CN111815375A (en) | User portrayal method and device in advertisement putting | |
| CN114268839A (en) | Video publishing account screening method and device, electronic equipment and storage medium | |
| CN113204699B (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN110188277B (en) | Resource recommendation method and device | |
| CN112989118A (en) | Video recall method and device | |
| CN112667890A (en) | Live broadcast interaction user behavior analysis method and device and computer equipment | |
| CN111523028A (en) | Data recommendation method, device, equipment and storage medium | |
| CN111314790A (en) | Video playing record sequencing method and device and electronic equipment | |
| CN110971973A (en) | Video pushing method and device and electronic equipment | |
| CN113408470B (en) | Data processing method, device, electronic equipment, storage medium and program product | |
| CN110874639A (en) | Method and device for acquiring operation information | |
| Werner et al. | Optimizing and Evaluating Stream-based News Recommendation Algorithms. | |
| CN111241821A (en) | Method and device for determining behavior characteristics of user | |
| CN113901261A (en) | Abnormal object cluster positioning method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210416 |