CN115269712A

CN115269712A - User interest mining method and system combined with meta-universe interaction service

Info

Publication number: CN115269712A
Application number: CN202210754502.8A
Authority: CN
Inventors: 李继光
Original assignee: Individual
Current assignee: Individual
Priority date: 2022-06-30
Filing date: 2022-06-30
Publication date: 2022-11-01

Abstract

The embodiment of the application discloses a user interest mining method and a user interest mining system combined with a metastic interaction service, which can enable a user interest mining algorithm to obtain digital space environment characteristics when mining and absorbing user interest, so that mining and identifying performances of the digital space environment characteristics are obtained, and mining, analyzing and absorbing interest events/interest knowledge of metastic user activity information in a first exemplary user activity information set and a second exemplary user activity information set along with an absorbing stage of the mining and identifying performances of the digital space environment characteristics, so that the user interest mining algorithm can accurately and credibly mine corresponding user interest events for metastic user activity information obtained under various digital space environment characteristics based on coping effects of the metastic user activity information of various digital space environment characteristics.

Description

User interest mining method and system combined with meta universe interaction service

Technical Field

The application relates to the technical field of meta universe, in particular to a user interest mining method and system combining meta universe interaction service.

Background

Along with the great heat of the concept of the metaseque, a plurality of digital enterprises are arranged in advance, and the metaseque technology and related industries are rapidly developed. Essentially, the metauniverse interaction is a novel digital virtualization space connecting a real world and a virtual world, which is obtained by further integrating and upgrading on the basis of big data, cloud computing and artificial intelligence. Currently, the user interaction analysis popularity for the meta universe is not reduced, various technologies favor the user big data mining in the meta universe scene, and most technologies cannot guarantee the precision and the credibility of interest mining for the user interest analysis of the meta universe service.

Disclosure of Invention

An object of the present application is to provide a user interest mining method and system in combination with a meta universe interaction service.

The technical scheme of the application is realized by at least some of the following embodiments.

A user interest mining method combined with a meta universe interaction service is applied to a meta universe interaction service system, and the method comprises the following steps: obtaining first meta-universe user activity information; and mining the user interest of the first universe user activity information based on a user interest mining algorithm to determine user interest knowledge distribution, wherein the user interest mining algorithm is configured by a first exemplary user activity information set and a second exemplary user activity information set, a reference indication corresponding to first authenticated user activity information in the first exemplary user activity information set matches a user interest event in the first authenticated user activity information, and a reference indication corresponding to second authenticated user activity information in the second exemplary user activity information set reflects digital spatial environment characteristics corresponding to the second authenticated user activity information.

The method is applied to relevant embodiments, user interest mining is carried out on the first meta-universe user activity information based on a user interest mining algorithm, the configuration of the user interest mining algorithm includes a first exemplary user activity information set and a second exemplary user activity information set, and second authenticated user activity information in the second exemplary user activity information set carries digital space environment characteristics, so that the user interest mining algorithm can obtain the digital space environment characteristics when carrying out user interest mining and absorption, so that mining and identification performances of the digital space environment characteristics are obtained, mining, analyzing and absorbing interest events/interest knowledge of the meta-universe user activity information in the first exemplary user activity information set and the second exemplary user activity information set along with an absorption stage of the mining and identification performances of the digital space environment characteristics, and accordingly the user interest mining algorithm can accurately and truthfully mine corresponding user interest events on the meta-universe user activity information obtained under various digital space environment characteristics.

Under some design considerations which can be independent, the user interest knowledge distribution contains digital space environment characteristics corresponding to the first metastic user activity information, and the method further comprises the following steps: and determining an activity attention variable corresponding to second meta-space user activity information through the digital space environment characteristics corresponding to the first meta-space user activity information, wherein the second meta-space user activity information is acquired after the first meta-space user activity information is acquired.

When the method is applied to related embodiments, the activity attention variable of the second meta-space user activity information can be determined by obtaining the digital space environment characteristics corresponding to the first meta-space user activity information, and a basis is further provided for user portrait sketching, meta-space user activity information analysis, subsequent processing of activity information and the like. For example, the active attention variable is used as an output of an active information crawler, so that an information collection thread performing collection based on the active information crawler can obtain the second binary cosmic user activity information based on the active attention variable, and for example, the active attention variable can also be used for subsequent processing of the activity information, and after the second binary cosmic user activity information is obtained, intelligent improvement of a heat value can be performed on the second binary cosmic user activity information based on the active attention variable. Alternatively, the active attention variable may also be used in the field of metastic security detection, based on which the above second metastic user activity information is analyzed.

Under some design considerations which can be independent, the user interest knowledge distribution comprises virtual community interest keywords, and the method further comprises the following steps: mining a user activity preference vector of the virtual community interest keyword; determining digital space interaction members corresponding to the interest keywords of the virtual community through the user activity preference vector; enabling a setup process for the digital space interaction member, the setup process including one or more of: portrait marking processing, target member liveness improvement processing and security authentication processing. The method is applied to the relevant embodiment, the extraction of the interest keywords of the virtual community can be completed by obtaining the user interest events in the first universe user activity information, the specific analysis processing can be carried out on the digital space interaction members of the interest keywords of the virtual community, and a basis is provided for part of requirements based on the extraction of the interest keywords of the virtual community.

Under some independent design ideas, the user interest knowledge distribution comprises virtual community interest keywords and digital space environment characteristics, and the method further comprises the following steps: and performing continuous analysis fusing the digital space environment characteristics on the interest keywords of the virtual community. The method is applied to the relevant embodiment, and continuity analysis can be carried out based on the digital space environment characteristics, so that the interference of the digital space environment characteristics on interest continuity analysis can be reduced, and the interest continuity analysis precision is improved to a certain extent.

Under some independent design considerations, the method further comprises: obtaining a first AI algorithm, wherein the first AI algorithm comprises a detail information extraction unit and a user interest mining unit, and the user interest mining unit is connected with the detail information extraction unit; deploying an environmental feature analysis unit and a support vector machine unit, wherein the environmental feature analysis unit is used for analyzing the digital space environmental features of the meta-cosmic user activity information loaded to the detail information extraction unit, the support vector machine unit is used for identifying an example information set corresponding to the meta-cosmic user activity information loaded to the detail information extraction unit, and the environmental feature analysis unit and the support vector machine unit are respectively connected with the detail information extraction unit in the first AI algorithm to determine a second AI algorithm; configuring the second AI algorithm with the first exemplary user activity information set and the second exemplary user activity information set to determine the user interest mining algorithm.

The user interest mining algorithm is applied to the related embodiments, the user interest mining algorithm is obtained based on the first AI algorithm, the timeliness of the user interest mining algorithm which tends to be stable can be improved based on the secondary use of the algorithm, the user interest mining algorithm in the embodiment of the application is obtained in time, and the user interest mining algorithm has the performance of analyzing the environmental features and mining the user interest based on the result of the environmental feature analysis.

Under some independent design ideas, before the environmental feature analyzing unit and the support vector machine unit are respectively connected to the detail information extracting unit in the first AI algorithm to determine the second AI algorithm, the method further includes: pre-configuring the first AI algorithm through the first exemplary user activity information set, improving the variable data of the detail information extraction unit and the user interest mining unit; the connecting the environmental feature analysis unit and the support vector machine unit with the detail information extraction unit in the first AI algorithm respectively to determine a second AI algorithm includes: and respectively connecting the environmental feature analysis unit and the support vector machine unit with a detail information extraction unit for finishing the improvement of variable data so as to determine the second AI algorithm.

As applied to the related embodiments, the first exemplary user activity information set can be used to configure a first AI algorithm and a second AI algorithm, which can be reused upon configuration, reducing the complexity of the acquisition upon which the configuration link is configured. The second AI algorithm is obtained based on the pre-configured first AI algorithm, so that the configuration efficiency of the second AI algorithm can be improved, and the timeliness of the model tending to stability is improved.

Under some independent design considerations, said configuring the second AI algorithm with the first exemplary user activity information set and the second exemplary user activity information set to determine the user interest mining algorithm includes: respectively mining the detail information of the first authenticated user activity information and the second authenticated user activity information through the detail information extraction unit to correspondingly obtain a first authenticated activity detail vector and a second authenticated activity detail vector; respectively carrying out interest event mining on the first authenticated activity detail vector and the second authenticated activity detail vector through the user interest mining unit to correspondingly obtain a first interest event mining result and a second interest event mining result; respectively carrying out authentication processing on the first authenticated activity detail vector and the second authenticated activity detail vector through the support vector machine unit to correspondingly obtain a first authentication processing result and a second authentication processing result; determining a first algorithm evaluation index according to the first interest event mining result, the first authentication processing result and the reference indication corresponding to the first authenticated user activity information; determining a second algorithm evaluation index according to the second interest event mining result, the second identification processing result and the reference indication corresponding to the second authenticated user activity information; and improving the variable data of the second AI algorithm through the first algorithm evaluation index and the second algorithm evaluation index to determine the user interest mining algorithm.

The method and the device are applied to relevant embodiments, the configuration of the second AI algorithm can be completed based on two types of configuration bases with different annotations to determine the user interest mining algorithm, the information in the two types of different configuration bases is absorbed and transfer learning is carried out, the configuration process has the performance of noise configuration by considering the category characteristics, and the mining precision of the user interest mining algorithm is improved.

In some independent design considerations, before determining a second algorithm evaluation index by using the second interest event mining result, the second identification processing result, and the reference indication corresponding to the second authenticated user activity information, the method further includes: performing digital space environment feature analysis on the second authenticated activity detail vector through the environment feature analysis unit to determine environment feature analysis data; the determining a second algorithm evaluation index according to the second interest event mining result, the second authentication processing result, and the reference indication corresponding to the second authenticated user activity information includes: and determining the second algorithm evaluation index through the second interest event mining result, the environmental feature analysis data, the second identification processing result and the reference indication corresponding to the second authenticated user activity information.

The method is applied to relevant embodiments, algorithm evaluation indexes output by the environment feature analysis unit, the support vector machine unit and the user interest mining unit can be effectively integrated in the process of determining the second algorithm evaluation index, so that the second algorithm evaluation index is obtained, the environment feature analysis unit, the support vector machine unit and the user interest mining unit are optimized based on the second algorithm evaluation index, variable data of the user interest mining unit can be free from interference of digital space environment features, and accurate and reliable interest event mining results can be obtained under various digital space environment features.

Under some independent design ideas, the determining the second algorithm evaluation index by using the second interest event mining result, the environmental feature analysis data, the second authentication processing result, and a reference indication corresponding to the second authenticated user activity information includes: determining an environmental characteristic analysis quality inspection index through a comparison result between the environmental characteristic analysis data and a reference indication corresponding to the second authenticated user activity information; determining an authentication quality indicator via a comparison between the second authentication processing result and a prior annotation corresponding to the second authenticated user activity information; determining derived indication data via the second interest event mining result, determining an interest mining quality indicator via the second interest event mining result and the derived indication data, the derived indication data defining a sort pattern of the second interest event mining result; analyzing the quality inspection indicators, the identification quality inspection indicators and the interest mining quality inspection indicators via the environmental characteristics to determine the second algorithm evaluation indicators.

As applied to the related embodiments, the second AI algorithm may be configured using the second exemplary set of user activity information as configuration indication information. The algorithm evaluation indexes generated by the environmental feature analysis unit, the support vector machine unit and the user interest mining unit on the premise of processing the second authenticated user activity information are obtained through accurate calculation, so that the timeliness of the second AI algorithm tending to stability in the configuration process can be accelerated, and the user interest mining accuracy of the configured user interest mining algorithm under various digital space environmental features is improved.

Under some independent design ideas, the second interest event mining result includes not less than one interest capture window, and a contribution coefficient corresponding to each interest capture window, where the contribution coefficient reflects a possibility that an interest event exists in a distribution area where the interest capture window is located, and the determining, by the second interest event mining result, derived indication data includes: for each interest capturing window, determining the interest capturing window as a first interest capturing window according to the condition that the contribution coefficient corresponding to the interest capturing window is higher than a first judgment value, and determining derivative indication data corresponding to the first interest capturing window as a first specified value, wherein the first specified value reflects that an interest event exists in a distribution area where the first interest capturing window is located; according to the condition that the contribution coefficient corresponding to the interest capture window is lower than a second judgment value, determining the interest capture window as a second interest capture window, and determining the derived indication data corresponding to the second interest capture window as a second judgment value, wherein the second judgment value reflects that no interest event exists in a distribution area where the second interest capture window is located.

The matching expected integration information can be determined on the basis of determining the sorting mode, so that variable data of the second AI algorithm can be improved on the basis of the expected integration information, and the mining identification performance of the user interest mining unit can be improved.

Under some independent design considerations, the determining an interest mining quality inspection indicator via the second interest event mining result and the derivative indication data includes: for each first interest capture window, determining a first interest capture evaluation index corresponding to the first interest capture window via a comparison result between the contribution coefficient corresponding to the first interest capture window and the first specified value; for each second interest capture window, determining a second interest capture evaluation index corresponding to the second interest capture window via a comparison result between the contribution coefficient corresponding to the second interest capture window and the second designated value; determining the interest mining quality metrics via the first and second interest capture evaluation metrics.

The method and the device are applied to relevant embodiments, the interest mining quality inspection index can be obtained only based on the first interest capturing evaluation index output by the first interest capturing window and the second interest capturing evaluation index output by the second interest capturing window, so that variable data in the second AI algorithm can be correspondingly sorted on the premise of definitely determining a sorting mode, errors are reduced, and the mining identification performance of the user interest mining unit is improved.

Under some independent design ideas, the second interest event mining result includes not less than one interest capture window, and a contribution coefficient corresponding to each interest capture window, where the contribution coefficient reflects a possibility that an interest event exists in a distribution area where the interest capture window is located, and the determining, by using the second interest event mining result, derivation indication data further includes: and determining the interest mining quality inspection index as a third specified value according to the condition that the contribution coefficient corresponding to each interest capturing window is smaller than the first judgment value and the contribution coefficient corresponding to each interest capturing window is higher than the second judgment value.

The method is applied to the related embodiment, the sorting confusion can be avoided, and under other conditions, the interest mining quality inspection indexes can be directly obtained, so that the operation complexity is reduced, the determination efficiency of the interest mining quality inspection indexes is improved, and the configuration process is accelerated.

Under some independent design ideas, the user interest knowledge distribution comprises virtual community interest keywords, and the mining of the user interest on the first universe user activity information based on the user interest mining algorithm to determine the user interest knowledge distribution comprises the following steps: performing detail information mining on the first user activity information of the first user in the universe through the detail information extraction unit to determine an activity detail array corresponding to the first user activity information of the first user in the universe; and mining the interest events of the activity detail array through the user interest mining unit to determine the interest keywords of the virtual community.

The method is applied to the relevant embodiments, the virtual community interest keywords can be timely output on the premise that the environment feature analysis unit of the user interest mining algorithm is not started, the identification efficiency of the virtual community interest keywords is improved, and the user interest mining unit of the user interest mining algorithm can obtain accurate and reliable user interest knowledge distribution under various digital space environment features including cold digital space environment features.

Under some independent design ideas, the user interest knowledge distribution further includes digital space environment features corresponding to the first metastic user activity information, and after the detail information extraction unit performs detail information mining on the first metastic user activity information to determine an activity detail array corresponding to the first metastic user activity information, the method further includes: and performing environmental characteristic analysis on the activity detail array through the environmental characteristic analysis unit to determine the digital space environmental characteristics corresponding to the first metacarpal user activity information.

The method is applied to the relevant embodiments, and the digital space environment characteristic analysis of the first universe user activity information can be realized through the environment characteristic analysis unit of the user interest mining algorithm so as to determine the accurate and reliable digital space environment characteristic.

A meta universe interaction service system comprising: a memory for storing an executable computer program, a processor for implementing the above method when executing the executable computer program stored in the memory.

A computer-readable storage medium, on which a computer program is stored which, when executed, performs the above-described method.

Drawings

FIG. 1 is a schematic diagram illustrating one communication configuration of a Meta-universe interaction service system in which embodiments of the present application may be implemented.

FIG. 2 is a flow diagram illustrating a method for user interest mining in conjunction with a meta universe interaction service in which embodiments of the application may be implemented.

FIG. 3 is an architectural diagram illustrating an application environment in which a user interest mining method in conjunction with a meta universe interaction service of embodiments of the present application may be implemented.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the attached drawings, the described embodiments should not be considered as limiting the present application, and all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application. In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the application.

FIG. 1 is a block diagram illustrating one communication configuration of a Meta-universe interaction service system 100 in which embodiments of the present application may be implemented, the Meta-universe interaction service system 100 including a memory 101 for storing an executable computer program, and a processor 102 for implementing a user interest mining method in conjunction with a Meta-universe interaction service in embodiments of the present application when executing the executable computer program stored in the memory 101.

Fig. 2 is a flowchart illustrating a user interest mining method in conjunction with a meta universe interaction service, which may implement an embodiment of the present application, and may be implemented by the meta universe interaction service system 100 illustrated in fig. 1, and further may include the technical solutions described in the following related steps.

And step 101, obtaining the first member universe user activity information.

In some examples, the first meta space user activity information is to-be-processed data of a user interest mining method combined with a meta space interaction service in the embodiment of the present application, and the embodiment of the present application does not limit an acquisition manner of the first meta space user activity information, and may be acquired by an information acquisition thread, or some group of component space user activity information in a meta space activity record, and further may also be derived from a shared database or other systems.

The embodiment of the application also does not limit the form of the meta-universe user activity information of the first meta-universe user activity information, for example, the form of the meta-universe user activity information may be a text form, an image-text form or other forms.

Further, the meta universe user activity information may include user activity information for a virtual webgame, user activity information for a VR/AR/3D projection, user activity information for a virtual community, user activity information for a virtual mall. Further, the meta universe user activity information may record a user's list of operational behavior data, interactivity data, and the like.

And 102, mining the user interest of the first universe user activity information based on a user interest mining algorithm to determine the user interest knowledge distribution.

Further, the user interest mining algorithm is configured by a first exemplary user activity information set and a second exemplary user activity information set, a reference indication corresponding to first authenticated user activity information in the first exemplary user activity information set points to a user interest event in the first authenticated user activity information, and a reference indication corresponding to second authenticated user activity information in the second exemplary user activity information set reflects a digital spatial environment characteristic corresponding to the second authenticated user activity information.

For example, the user interest knowledge distribution may be understood as a record set corresponding to the interest knowledge feature, and thus may be understood as an interest knowledge feature map, and the user interest event includes a series of preferences and preference items of the user in the meta universe interaction service, such as a rendering preference for VR/AR/3D projection, a somatosensory optimization requirement for a virtual game, a commodity recommendation preference for a virtual mall, and the like, which are not limited herein. On this basis, the digital space environment features are used for reflecting the interaction environment of the meta universe service, such as the interaction scale, the regional distribution of the participating users, the network security condition, and the like, and are not limited herein.

According to the embodiment of the application, the user interest mining algorithm is used for mining the user interest of the first meta-universe user activity information, the configuration basis of the user interest mining algorithm comprises a first exemplary user activity information set and a second exemplary user activity information set, and the second authenticated user activity information in the second exemplary user activity information set carries digital space environment characteristics, so that the user interest mining algorithm can obtain the digital space environment characteristics when mining and absorbing the user interest, so that the mining and identification performance of the digital space environment characteristics is obtained, and mining, analyzing and absorbing interest events/interest knowledge of the meta-universe user activity information in the first exemplary user activity information set and the second exemplary user activity information set along with an absorption stage (learning stage) of the mining and identification performance of the digital space environment characteristics, so that the user interest mining algorithm can accurately mine the corresponding user interest events in a universe credible manner on the meta-universe user activity information obtained under various digital space environment characteristics.

In some examples, the user interest knowledge distribution includes a digital space environment characteristic corresponding to the first meta universe user activity information. Based on the user interest knowledge distribution, corresponding processing analysis may be performed, which is not further limited in the embodiment of the present application. For example, an activity attention variable (a parameter for reflecting the weight/importance degree of different activity information) corresponding to second meta-space user activity information, which is collected after the first meta-space user activity information is obtained, may be determined according to the digital space environment characteristics corresponding to the first meta-space user activity information.

Further, the active attention variable may be applied to a plurality of application fields, for example, the active attention variable may be used as an output result of an active information crawler, so that an information collecting thread collecting based on the active information crawler may collect the second binary cosmic user activity information based on the active attention variable. As another example, the active attention variable may also be used for subsequent processing of the activity information, and after obtaining the second cosmic user activity information, intelligent improvements to the thermal value may be made to the second cosmic user activity information based on the active attention variable. For another example, the active attention variable may be used in the field of metastic security detection, and the second metastic user activity information is analyzed and processed based on the active attention variable. According to the content, the activity attention variable of the second meta-space user activity information can be determined by obtaining the digital space environment characteristics corresponding to the first meta-space user activity information, and a basis is further provided for user portrait sketching, meta-space user activity information analysis, subsequent processing of activity information and the like.

In some examples, the user interest knowledge distribution further includes a virtual community interest keyword based on which corresponding processing analysis may be performed, and in some examples, a user activity preference vector (which may be understood as a user activity preference feature) of the virtual community interest keyword may be mined; determining digital space interaction members (such as virtual user members) corresponding to the virtual community interest keywords based on the user activity preference vector; enabling a setup process for the digital space interaction member, the setup process including one or more of: the method comprises the steps of image marking processing, target member activity improvement processing and security authentication processing.

For example, taking the virtual community interest keyword as a community group purchase as an example, a user activity preference vector of the community group purchase may be mined, for example: and if the comparison analysis is completed, the digital space interactive member pointed by the user activity preference vector determined by the comparison analysis is used as the digital space interactive member of the virtual community interest keyword, so that the analysis of the virtual community interest keyword is completed. On the basis of obtaining the digital space interactive member of the virtual community interest keyword, portrait marking processing, state adjustment processing, security authentication processing and the like of the digital space interactive member can be performed, which is not further limited in the embodiment of the present application. According to the content, the analysis of the virtual community interest keywords can be completed through the user interest events in the mined first universe user activity information, and the digital space interactive members of the virtual community interest keywords can be subjected to targeted analysis processing, so that a basis is provided for part of analysis requirements based on the virtual community interest keywords.

On the premise that the user interest knowledge distribution comprises virtual community interest keywords and digital space environment characteristics, continuous analysis based on the digital space environment characteristics can be carried out on the virtual community interest keywords. The interest persistence analysis can be used for carrying out real-time user interest processing, so that tracking mining processing of user big data is achieved, and in the embodiment of the application, persistence analysis can be carried out based on digital space environment characteristics, so that interference of the digital space environment characteristics on the interest persistence analysis can be reduced, and the interest persistence analysis accuracy is improved to a certain extent.

In a possible embodiment, which shows the configuration idea of the user interest mining algorithm, the method may further include the recorded contents of steps 201 to 204.

Step 201, obtaining a first AI algorithm, where the first AI algorithm includes a detail information extraction unit and a user interest mining unit, and the user interest mining unit is connected to the detail information extraction unit.

In the embodiment of the present application, the detail information extraction unit may be understood as a feature extractor, and the user interest mining unit may be understood as a user interest identifier.

The first AI algorithm is not limited in the embodiment of the application, can be an algorithm with a user interest mining function configured in advance, can be an algorithm directly used for mining the user interest in the related technology, is obtained based on the prior first AI algorithm, can improve the stability timeliness of the user interest mining algorithm based on the secondary use of the algorithm, and can be obtained in time.

Further, the first AI algorithm is not limited in this embodiment, and may be a convolutional neural network, a deep learning network, a cyclic neural network, a multi-layer perceptron, a naive bayesian network, or the like.

Step 202, deploying an environmental feature analysis unit and a support vector machine unit, wherein the environmental feature analysis unit is used for analyzing the digital space environmental features of the meta-universe user activity information loaded to the detail information extraction unit, and the support vector machine unit is used for identifying an example information set corresponding to the meta-universe user activity information loaded to the detail information extraction unit.

Furthermore, the environmental feature analysis unit may also be understood as an environmental feature detector. The support vector machine unit may also be understood as a classifier. Deploying the environment feature parsing unit and the support vector machine unit may be understood as constructing the environment feature parsing unit and the support vector machine unit.

In the configuration link of the user interest mining algorithm, contents related to digital space environment feature mining and contents related to user interest mining, in other words, contents related to two different application environments need to be combined to complete user interest mining, so that derivative analysis processing of the two types of information needs to be performed in the model in the configuration link, and the derivative analysis processing (transfer learning) can be realized based on a support vector machine unit.

Step 203, the environmental feature analysis unit and the support vector machine unit are respectively connected with the detail information extraction unit in the first AI algorithm to determine a second AI algorithm.

In the embodiment of the application, the second AI algorithm can be obtained by performing algorithm configuration optimization on the basis of the first AI algorithm, and the related introduction of the second AI algorithm is as follows. Further, the user interest mining unit _2, the environmental feature analysis unit _3, and the support vector machine unit _4 are all connected to the detail information extraction unit _1, in other words, the content derived by the detail information extraction unit _1 can be regarded as raw material information of the user interest mining unit _2, the environmental feature analysis unit _3 and the support vector machine unit _4, so that the user interest mining algorithm configured based on the second AI algorithm has the functions of environmental feature analysis and user interest mining based on the result of the environmental feature analysis.

Another algorithm architecture of the second AI algorithm of the embodiment of the present application may further include the following contents. The main algorithm architecture is used as a part of the detail information extraction unit and can be used for carrying out basic detail information mining on the imported meta-universe user activity information. The imported metauniverse user activity information may be derived from the first exemplary set of user activity information and/or the second exemplary set of user activity information.

The embodiments of the present application do not limit the main algorithm architecture, for example: may be the basic CNN. The gradient algorithm is connected with the main algorithm framework and also serves as a part of the detail information extraction unit, and can be used for mining multi-dimensional activity detail arrays, so that the activity details in the meta-universe user activity information can be more completely obtained. The content derived by the gradient algorithm can be led into a user interest mining unit, the user interest mining unit can estimate the state and type information of the interest event, the type information reflects the attribute of the interest event, and the type information reflects the possibility of community group buying when the interest event is mined out by taking community group buying mining as an example. The state of the interest event is reflected through the positioning node, and the type information of the interest event is reflected through the classification node. The content derived by the gradient algorithm can be input into an environment feature analysis unit and a support vector machine unit so as to respectively evaluate the digital space environment feature and the activity thermodynamic distribution of the meta-universe user activity information, and the digital space environment feature and the activity thermodynamic distribution of the meta-universe user activity information are reflected through a dynamic thermodynamic diagram. Processing of the meta universe user activity information derived from the first exemplary set of user activity information may be recorded via a first flow trace and processing of the meta universe user activity information derived from the second exemplary set of user activity information may be recorded via a second flow trace. In actual implementation, one interest event is reflected by one capture unit, one capture unit corresponds to one interest capture window, and each interest capture window has state, category, digital space environment characteristics and activity thermodynamic data.

Step 204, configuring the second AI algorithm based on the first exemplary user activity information set and the second exemplary user activity information set to determine the user interest mining algorithm.

In the embodiment of the present application, the first authenticated user activity information of the first exemplary user activity information set all covers the reference indication (label information) corresponding to the interest event, for example: annotation information for community group purchases may be covered. The second authenticated user activity information of the second exemplary user activity information set covers the digital space environment feature tag, and the first authenticated user activity information and the second authenticated user activity information may be different metastic user activity information or the same metastic user activity information, which is not further limited by the embodiment of the present application. It will be appreciated that it is relatively convenient to obtain a reference indication corresponding to an event of interest or a tag of a digital spatial environment characteristic independently, which enables a flexible and efficient determination of an example information set.

In the embodiment of the application, the first AI algorithm may be configured in advance based on the first exemplary user activity information set, and the variable data of the detail information extraction unit and the user interest mining unit are improved, in other words, the first exemplary user activity information set may be used to configure the first AI algorithm and the second AI algorithm, which may be used repeatedly, so as to reduce the complexity of obtaining configuration bases of configuration links. And respectively connecting the environmental feature analysis unit and the support vector machine unit with a detail information extraction unit for finishing the improvement of variable data to determine the second AI algorithm. The second AI algorithm is obtained based on the pre-configured first AI algorithm, so that the configuration efficiency of the second AI algorithm can be improved, and the timeliness of the model tending to be stable is improved.

In one possible embodiment, the configuring the second AI algorithm to determine the user interest mining algorithm based on the first exemplary user activity information set and the second exemplary user activity information set may include recording as follows from steps 301-305.

Step 301, based on the detail information extraction unit, performing detail information mining on the first authenticated user activity information and the second authenticated user activity information respectively, and correspondingly obtaining a first authenticated activity detail vector and a second authenticated activity detail vector.

By way of example, an authenticated activity detail vector may be understood as a sample feature or sample feature vector for feature data reflecting sample activity information.

Step 302, respectively performing interest event mining on the first authenticated activity detail vector and the second authenticated activity detail vector based on the user interest mining unit, and correspondingly obtaining a first interest event mining result and a second interest event mining result.

Step 303, performing authentication processing on the first authenticated activity detail vector and the second authenticated activity detail vector respectively based on the support vector machine unit, and correspondingly obtaining a first authentication processing result and a second authentication processing result.

The identification process is understood to be a classification process.

Step 304, determining a first algorithm evaluation index based on the first interest event mining result, the first authentication processing result and the reference indication corresponding to the first authenticated user activity information.

For example, the algorithm evaluation index output by the user interest mining unit may be obtained according to a comparison result between the first interest event mining result and the reference indication corresponding to the first authenticated user activity information, and the algorithm evaluation index output by the support vector machine unit may be directly obtained according to the first authentication processing result, for example, the first authenticated user activity information is determined to be derived from the first exemplary user activity information set, but the first authentication processing result reflects that the probability that the first authenticated user activity information is derived from the first exemplary user activity information set is only 0.8, and the support vector machine unit generates the algorithm evaluation index (loss). The two algorithm evaluation indexes are subjected to global processing (weighting processing) to obtain a first algorithm evaluation index, the value of a global processing result is not limited in the embodiment of the application, and flexible setting can be performed based on actual needs. The algorithm evaluation index output by the user interest mining unit can be recorded through the hinge algorithm evaluation index. The algorithm evaluation index output by the support vector machine unit can be recorded through a linkage algorithm evaluation index. Further, the hinge algorithm evaluation index can be understood as a hinge loss function, and the linkage algorithm evaluation index can be understood as a countermeasure loss function.

Step 305, determining a second algorithm evaluation index based on the second interest event mining result, the second authentication processing result and the reference indication corresponding to the second authenticated user activity information.

In this embodiment of the present application, digital spatial environment feature analysis may be further performed on the second authenticated activity detail vector based on the environment feature analysis unit to determine environment feature analysis data. And determining the second algorithm evaluation index according to the second interest event mining result, the environment feature analysis data, the second identification processing result and the reference indication corresponding to the second authenticated user activity information. In the process of determining the second algorithm evaluation index, the algorithm evaluation indexes output by the environment feature analysis unit, the support vector machine unit and the user interest mining unit can be effectively integrated, so that the second algorithm evaluation index is obtained, the environment feature analysis unit, the support vector machine unit and the user interest mining unit are optimized based on the second algorithm evaluation index, the variable data of the user interest mining unit can be free from the interference of the digital space environment features, and accurate and reliable interest event mining results can be obtained under various digital space environment features.

In a possible embodiment, the determining a second algorithm evaluation indicator based on the second interest event mining result, the second authentication processing result, and the reference indication corresponding to the second authenticated user activity information may include the following: and determining an environmental characteristic analysis quality inspection index according to a comparison result between the environmental characteristic analysis data and a reference indication corresponding to the second authenticated user activity information. And determining an authentication quality index according to a comparison result between the second authentication processing result and a prior annotation (actual source) corresponding to the second authenticated user activity information. And determining derived indication data according to the second interest event mining result, and determining an interest mining quality inspection index according to the second interest event mining result and the derived indication data, wherein the derived indication data is used for limiting a sorting mode (an integration method) of the second interest event mining result. And analyzing the quality inspection indexes, the identification quality inspection indexes and the interest mining quality inspection indexes according to the environmental characteristics to determine the second algorithm evaluation indexes.

As applied to the related embodiments, the second AI algorithm may be configured using the second exemplary set of user activity information as the configuration indication information. The algorithm evaluation indexes generated by the environmental feature analysis unit, the support vector machine unit and the user interest mining unit on the premise of processing the second authenticated user activity information are obtained through accurate calculation, so that the timeliness of the second AI algorithm which tends to be stable in the configuration process can be accelerated, and the user interest mining accuracy of the configured user interest mining algorithm under various digital space environmental features is improved.

In the embodiment of the application, the environmental characteristic analysis quality inspection index can also be expressed by using a hinge algorithm evaluation index, and the identification quality inspection index can be described by using a linkage algorithm evaluation index. The second algorithm evaluation index may be a global processing result of the environmental characteristic analysis quality inspection index, the identification quality inspection index, and the interest mining quality inspection index, and of course, the value of the global processing result is not limited in the embodiment of the present application, and may be flexibly set based on actual needs.

Further, the quality control index described above can be understood as a prediction loss in the related object mining process.

In a possible embodiment, the second interest event mining result includes no less than one interest capture window, and a contribution coefficient corresponding to each interest capture window, where the contribution coefficient reflects a possibility that an interest event exists in a distribution area where the interest capture window is located, and the determining of the derivative indication data according to the second interest event mining result includes steps 401 and 402.

Step 401, for each interest capture window (such as a convolution kernel), according to a condition that a contribution coefficient corresponding to the interest capture window is higher than a first determination value, determining the interest capture window as a first interest capture window, and determining derivative indication data corresponding to the first interest capture window as a first specified value, where the first specified value reflects that an interest event exists in a distribution area where the first interest capture window is located.

For example, for the interest capture window _ a, the corresponding contribution coefficient (confidence) is 0.98, and the first decision value is 0.9, in other words, the possibility that the community group purchase exists in the interest capture window _ a is larger than a first determination value, the community group purchase is highly likely to exist in the interest capture window _ a, in view of this, the contribution coefficient corresponding to the interest capture window _ a may be integrated to a first specified value reflecting the expected contribution coefficient that actually matches with the community group purchase situation, e.g., the first specified value may be 1.

Step 402, according to a condition that the contribution coefficient corresponding to the interest capture window is lower than a second determination value, determining the interest capture window as a second interest capture window, and determining the derived indication data corresponding to the second interest capture window as a second determination value, where the second determination value reflects that no interest event exists in a distribution area where the second interest capture window is located.

For example, for the interest capture window _ B, the corresponding contribution factor (such as confidence level) is 0.08, and the first determination value is 0.1, in other words, the probability of the community group purchase existing in the interest capture window _ B is lower than the second determination value, for example, the contribution coefficient corresponding to the interest capture window _ B may be integrated into a second specified value, and the second specified value reflects an expected contribution coefficient corresponding to the condition that there is no community group purchase, for example: the first specified value may be 0.

It is understood that, in the embodiment of the present application, the first determination value, the second determination value, the first specified value, and the second specified value may be flexibly set, and the present application does not further limit this. Applied to the related embodiment, the matching expected integration information can be determined on the basis of the determination of the sorting mode, so that the variable data of the second AI algorithm can be improved based on the expected integration information, and the mining identification performance of the user interest mining unit can be realized.

In a possible embodiment, the determining an interest mining quality index according to the second interest event mining result and the derived indication data may include the following: for each first interest capture window, determining a first interest capture evaluation index corresponding to the first interest capture window according to a comparison result between the contribution coefficient corresponding to the first interest capture window and the first designated value; for each second interest capture window, determining a second interest capture evaluation index corresponding to the second interest capture window according to a comparison result between the contribution coefficient corresponding to the second interest capture window and the second specified value; and determining the interest mining quality inspection index according to the first interest capturing evaluation index and the second interest capturing evaluation index.

The method is applied to the related embodiment, the interest mining quality inspection index can be obtained only based on the first interest capturing evaluation index output by the first interest capturing window and the second interest capturing evaluation index output by the second interest capturing window, so that variable data in the second AI algorithm can be correspondingly sorted on the premise of definitely determining the sorting mode, errors are reduced, and the mining identification performance of the user interest mining unit is improved. In the embodiment of the application, both the first interest capturing evaluation index and the second interest capturing evaluation index can be recorded by using a hinge algorithm evaluation index, and the interest mining quality inspection index can be a global processing result of the two types of interest capturing evaluation indexes.

In some examples, the interest mining quality inspection indicator is determined to be a third specified value according to a case that the contribution coefficient corresponding to each interest capture window is smaller than the first determination value and the contribution coefficient corresponding to each interest capture window is higher than the second determination value. The method is applied to the related embodiment, the sorting confusion can be avoided, and under other conditions, the interest mining quality inspection indexes are directly obtained, so that the operation complexity is reduced, the determination efficiency of the interest mining quality inspection indexes is improved, and the configuration process is accelerated. The value of the third specified value is not limited in this application, and may be set to "0", for example.

Step 306, modifying the variable data of the second AI algorithm based on the first algorithm evaluation index and the second algorithm evaluation index to determine the user interest mining algorithm.

In the embodiment of the application, the algorithm variable data of the second AI algorithm may be improved based on the first algorithm evaluation index and the second algorithm evaluation index, for example, the variable data may be improved by an optimal algorithm (for example, a gradient descent method) until the variable data meets a termination index, for example, the variable data may be terminated after the set number of rounds is performed in a loop, or the variable data may be terminated on the premise that both the first algorithm evaluation index and the second algorithm evaluation index are smaller than a preset termination judgment value.

In the embodiment of the application, for the first authenticated user activity information in the first authenticated user activity information set, the algorithm evaluation index output by the user interest mining unit may be obtained, where the algorithm evaluation index includes a distribution characteristic evaluation index and an information category evaluation index, and may also be obtained as a quality inspection identification index. For the second authenticated user activity information in the second authenticated user activity information set, an algorithm evaluation index output by the user interest mining unit can be obtained, the algorithm evaluation index comprises a distribution characteristic evaluation index and an information category evaluation index, and an identification quality inspection index and an environmental characteristic analysis quality inspection index can also be obtained. The method and the device are applied to relevant embodiments, the configuration of the second AI algorithm can be completed based on two types of configuration bases with different annotations to determine the user interest mining algorithm, the information in the two types of different configuration bases is absorbed and transfer learning is carried out, the configuration process has the performance of noise configuration by considering the category characteristics, and the mining precision of the user interest mining algorithm is improved.

Further, on the premise that a user interest mining algorithm is configured, detail information mining can be performed on the first user activity information of the first user in the universe based on the detail information extraction unit to determine an activity detail array corresponding to the first user activity information of the first user in the universe; and performing interest event mining on the activity detail array based on the user interest mining unit to determine the virtual community interest keywords. On the basis, the virtual community interest keywords can be timely output on the premise of not starting the environment feature analysis unit of the user interest mining algorithm, the identification efficiency of the virtual community interest keywords is improved, and the user interest mining unit of the user interest mining algorithm can obtain accurate and reliable user interest knowledge distribution under various digital space environment features including cold digital space environment features.

If the digital space environment characteristics of the first metacosmic user activity information need to be obtained, after the detail information mining is performed on the first metacosmic user activity information based on the detail information extraction unit to determine an activity detail array corresponding to the first metacosmic user activity information, environment characteristic analysis is performed on the activity detail array based on the environment characteristic analysis unit to determine the digital space environment characteristics corresponding to the first metacosmic user activity information. On the basis, the digital space environment characteristic analysis of the first universe user activity information can be realized through an environment characteristic analysis unit of a user interest mining algorithm so as to determine accurate digital space environment characteristics.

Under some independently implementable design considerations, after determining the user interest knowledge distribution, the method may further include: responding to a pushing auxiliary analysis request of a meta universe service platform system, and mining pushing requirements of the user interest knowledge distribution to obtain a user pushing requirement relation network corresponding to the user interest knowledge distribution; and sending the user pushing demand relation network to the meta universe service platform system.

The method is applied to the design idea, the pushing requirement mining is executed only when the pushing auxiliary analysis request is received, so that meaningless resource waste is avoided, the pushing requirement mining is carried out on the side of the metas-universe interaction service system, and the user interest knowledge distribution does not need to be sent to the metas-universe service platform system, so that the user privacy safety corresponding to the user interest knowledge distribution is guaranteed, and further, the pushing requirement relation network is sent to the metas-universe service platform system, so that the metas-universe service platform system can be directly guided to carry out targeted big data recommendation.

Under some design ideas which can be independently implemented, carrying out pushing demand mining on the user interest knowledge distribution to obtain a user pushing demand relation network corresponding to the user interest knowledge distribution, and realizing the pushing demand relation network by the following technical scheme: mining an interest node description field and an interest scene description field in user interest knowledge distribution; based on the description field correlation degree between the interest node description field and the interest scene description field in the user interest knowledge distribution, associating the interest node description field and the interest scene description field in the user interest knowledge distribution to obtain a description field association result; determining an interest scene description field which is not associated completely as a candidate interest scene description field, and determining a pushing requirement matched with the candidate interest scene description field according to a description field vector distance between the interest scene description field in the description field association result and the candidate interest scene description field; associating the pushing requirement matched with the candidate interest scene description field to obtain a requirement association result; and determining a user pushing demand relation network corresponding to the user interest knowledge distribution according to the demand association result and the description field association result.

For example, the interest node and the interest scene respectively correspond to local and global interest knowledge, so that correlation analysis of the push requirement can be realized by considering the description fields corresponding to the interest node and the interest scene, and a user push requirement relation network is accurately and completely mined.

Under some design ideas which can be independently implemented, the obtaining of the interest node description field and the interest scene description field in the user interest knowledge distribution includes: acquiring not less than two interest node data and not less than two interest scene data in the user interest knowledge distribution; acquiring first feature similarity and first feature influence degree between the at least two interest node data, and acquiring second feature similarity and second feature influence degree between the at least two interest scene data; splicing the data of the at least two interest nodes according to the first feature similarity and the first feature influence degree to obtain an interest node description field in the user interest knowledge distribution; an interest node description field includes at least one interest node data; splicing the not less than two pieces of interest scene data according to the second feature similarity and the second feature influence degree to obtain an interest scene description field in the user interest knowledge distribution; an interest scene description field includes at least one interest scene data.

By the design, the interest node description field and the interest scene description field can be completely determined, and the lack of the interest node description field and the interest scene description field is avoided.

Fig. 3 is an architecture diagram illustrating an application environment of a user interest mining method in conjunction with a meta universe interaction service, in which a meta universe interaction service system 100 and a meta universe interaction terminal 200, which communicate with each other, may be included, in which an embodiment of the present application may be implemented. Based on this, the metascosmic interaction service system 100 and the metascosmic interaction terminal 200 implement or partially implement the user interest mining method incorporating the metascosmic interaction service according to the embodiment of the present application at runtime.

The above description is only a preferred embodiment of the present application, and is not intended to limit the scope of the present application.

Claims

1. A user interest mining method combined with a meta-universe interaction service is applied to a meta-universe interaction service system, and the method comprises the following steps:

acquiring first user activity information of the universe, and mining user interest of the first user activity information of the universe based on a user interest mining algorithm to determine user interest knowledge distribution; the user interest mining algorithm is configured through a first exemplary user activity information set and a second exemplary user activity information set, a reference indication corresponding to first authenticated user activity information in the first exemplary user activity information set matches a user interest event in the first authenticated user activity information, and a reference indication corresponding to second authenticated user activity information in the second exemplary user activity information set reflects a digital spatial environment characteristic corresponding to the second authenticated user activity information.

2. The method of claim 1, wherein the user interest knowledge distribution includes digital space environment characteristics corresponding to the first metauniverse user activity information, the method further comprising:

and determining an activity attention variable corresponding to second meta-space user activity information through the digital space environment characteristics corresponding to the first meta-space user activity information, wherein the second meta-space user activity information is acquired after the first meta-space user activity information is acquired.

3. The method of claim 1, wherein the user interest knowledge distribution comprises virtual community interest keywords, the method further comprising:

mining a user activity preference vector of the virtual community interest keyword;

determining digital space interaction members corresponding to the interest keywords of the virtual community through the user activity preference vector;

enabling a setup process for the digital space interaction member, the setup process including one or more of: the method comprises the steps of image marking processing, target member activity improvement processing and security authentication processing.

4. The method of claim 1, further comprising:

obtaining a first AI algorithm, wherein the first AI algorithm comprises a detail information extraction unit and a user interest mining unit, and the user interest mining unit is connected with the detail information extraction unit;

deploying an environmental feature analysis unit and a support vector machine unit, wherein the environmental feature analysis unit is used for analyzing the digital spatial environmental features of the meta cosmic user activity information loaded to the detail information extraction unit, the support vector machine unit is used for identifying an example information set corresponding to the meta cosmic user activity information loaded to the detail information extraction unit, and the environmental feature analysis unit and the support vector machine unit are respectively connected with the detail information extraction unit in the first AI algorithm to determine a second AI algorithm;

configuring the second AI algorithm with the first exemplary user activity information set and the second exemplary user activity information set to determine the user interest mining algorithm.

5. The method according to claim 4, wherein before the connecting the environmental feature analysis unit and the support vector machine unit to the detail information extraction unit in the first AI algorithm to determine the second AI algorithm, the method further comprises: preconfiguring the first AI algorithm through the first exemplary user activity information set to improve variable data of the detail information extraction unit and the user interest mining unit;

the connecting the environmental feature analysis unit and the support vector machine unit with the detail information extraction unit in the first AI algorithm respectively to determine a second AI algorithm includes: and respectively connecting the environmental feature analysis unit and the support vector machine unit with a detail information extraction unit for finishing the improvement of variable data so as to determine the second AI algorithm.

6. The method of claim 4, wherein said configuring the second AI algorithm with the first exemplary user activity information set and the second exemplary user activity information set to determine the user interest mining algorithm comprises:

respectively mining detail information of the first authenticated user activity information and the second authenticated user activity information through the detail information extraction unit to correspondingly obtain a first authenticated activity detail vector and a second authenticated activity detail vector;

respectively carrying out interest event mining on the first authenticated activity detail vector and the second authenticated activity detail vector through the user interest mining unit to correspondingly obtain a first interest event mining result and a second interest event mining result;

respectively carrying out authentication processing on the first authenticated activity detail vector and the second authenticated activity detail vector through the support vector machine unit to correspondingly obtain a first authentication processing result and a second authentication processing result;

determining a first algorithm evaluation index according to the first interest event mining result, the first authentication processing result and the reference indication corresponding to the first authenticated user activity information;

determining a second algorithm evaluation index according to the second interest event mining result, the second identification processing result and the reference indication corresponding to the second authenticated user activity information;

improving the variable data of the second AI algorithm through the first algorithm evaluation index and the second algorithm evaluation index to determine the user interest mining algorithm;

before determining a second algorithm evaluation index according to the second interest event mining result, the second authentication processing result, and the reference indication corresponding to the second authenticated user activity information, the method further includes: performing digital space environment feature analysis on the second authenticated activity detail vector through the environment feature analysis unit to determine environment feature analysis data;

the determining a second algorithm evaluation index according to the second interest event mining result, the second authentication processing result, and the reference indication corresponding to the second authenticated user activity information includes: determining the second algorithm evaluation index through the second interest event mining result, the environmental feature analysis data, the second identification processing result and a reference indication corresponding to the second authenticated user activity information;

wherein the determining the second algorithm evaluation index by the second interest event mining result, the environmental feature analysis data, the second authentication processing result, and the reference indication corresponding to the second authenticated user activity information includes:

determining an environmental characteristic analysis quality inspection index through a comparison result between the environmental characteristic analysis data and a reference indication corresponding to the second authenticated user activity information;

determining an authentication quality indicator via a comparison between the second authentication processing result and a prior annotation corresponding to the second authenticated user activity information;

determining derived indication data via the second interest event mining result, determining an interest mining quality indicator via the second interest event mining result and the derived indication data, the derived indication data defining a sort pattern of the second interest event mining result;

analyzing the quality inspection indicators, the identification quality inspection indicators and the interest mining quality inspection indicators via the environmental characteristics to determine the second algorithm evaluation indicators.

7. The method of claim 6, wherein the second interest event mining results comprise no less than one interest capture window, and a contribution coefficient corresponding to each interest capture window, the contribution coefficient reflecting a probability that an interest event exists in a distribution area where the interest capture window is located, and wherein determining, via the second interest event mining results, derivation indication data comprises:

for each interest capture window, determining the interest capture window as a first interest capture window according to the condition that the contribution coefficient corresponding to the interest capture window is higher than a first judgment value, and determining derived indication data corresponding to the first interest capture window as a first specified value, wherein the first specified value reflects that an interest event exists in a distribution area where the first interest capture window is located;

determining the interest capture window as a second interest capture window according to the condition that the contribution coefficient corresponding to the interest capture window is lower than a second determination value, and determining derived indication data corresponding to the second interest capture window as a second determination value, wherein the second determination value reflects that no interest event exists in a distribution area where the second interest capture window is located;

wherein the determining an interest mining quality indicator via the second interest event mining result and the derived indication data comprises: for each first interest capture window, determining a first interest capture evaluation index corresponding to the first interest capture window via a comparison result between the contribution coefficient corresponding to the first interest capture window and the first specified value; for each second interest capture window, determining a second interest capture evaluation index corresponding to the second interest capture window via a comparison result between the contribution coefficient corresponding to the second interest capture window and the second designated value; determining the interest mining quality metrics via the first interest capture evaluation metric and the second interest capture evaluation metric;

wherein the second interest event mining result includes not less than one interest capture window and a contribution coefficient corresponding to each interest capture window, the contribution coefficient reflects a possibility that an interest event exists in a distribution area where the interest capture window is located, and the determining of the derivative indication data through the second interest event mining result further includes: and determining the interest mining quality inspection index as a third specified value according to the condition that the contribution coefficient corresponding to each interest capturing window is smaller than the first judgment value and the contribution coefficient corresponding to each interest capturing window is higher than the second judgment value.

8. The method of claim 4, wherein the user interest knowledge distribution includes virtual community interest keywords, and wherein the mining user interest from the first universe of user activity information based on a user interest mining algorithm to determine a user interest knowledge distribution comprises:

performing detail information mining on the first user activity information of the first user in the universe through the detail information extraction unit to determine an activity detail array corresponding to the first user activity information of the first user in the universe; and mining the interest events of the activity detail array through the user interest mining unit to determine the virtual community interest keywords.

9. The method of claim 4, wherein the user interest knowledge distribution further includes digital space environment features corresponding to the first metauniverse user activity information, and wherein after the detail information extraction unit performs detail information mining on the first metauniverse user activity information to determine an activity detail array corresponding to the first metauniverse user activity information, the method further comprises: and performing environmental characteristic analysis on the activity detail array through the environmental characteristic analysis unit to determine the digital space environmental characteristics corresponding to the first metacarpal user activity information.

10. A meta-universe interaction service system, comprising:

a memory for storing an executable computer program, a processor for implementing the method of any one of claims 1-9 when executing the executable computer program stored in the memory.