CN115712657A - User demand mining method and system based on meta universe - Google Patents
User demand mining method and system based on meta universe Download PDFInfo
- Publication number
- CN115712657A CN115712657A CN202211283432.9A CN202211283432A CN115712657A CN 115712657 A CN115712657 A CN 115712657A CN 202211283432 A CN202211283432 A CN 202211283432A CN 115712657 A CN115712657 A CN 115712657A
- Authority
- CN
- China
- Prior art keywords
- behavior habit
- user
- target
- requirement
- vector
- 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- Withdrawn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
- 238000005065 mining Methods 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 34
- 239000013598 vector Substances 0.000 claims abstract description 203
- 230000000694 effects Effects 0.000 claims abstract description 124
- 230000006399 behavior Effects 0.000 claims description 209
- 230000000007 visual effect Effects 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000013461 design Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012800 visualization Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
According to the user demand mining method and system based on the metauniverse, after the conversation behavior habit vector and the stage behavior habit vector of the metauniverse user activity big data are obtained, the conversation scene and the conversation link of the metauniverse user activity big data are in contact by conducting demand prediction on the conversation behavior habit vector and the stage behavior habit vector, so that when the demand theme mining is conducted on the metauniverse user activity big data based on the obtained target conversation behavior habit vector and the target stage behavior habit vector, a user demand theme which is as accurate and credible as possible can be obtained, and the quality of conducting the user demand mining on the metauniverse user activity big data is improved.
Description
Technical Field
The application relates to the technical field of a meta universe, in particular to a user demand mining method and system based on the meta universe.
Background
The metauniverse (Metaverse) is a digital living space which is constructed by applying digital technology, is a virtual world which is mapped or surpassed by the real world and can interact with the real world, and has a novel social system. Practical landing applications of the meta universe include, but are not limited to, virtual malls, digital communities, and the like. With the development of the metauniverse, the demand analysis for metauniverse users is of great importance, but the traditional technology is difficult to guarantee the quality of the demand analysis/mining.
Disclosure of Invention
One object of the present application is to provide a user demand mining method and system based on the metauniverse.
A user demand mining method based on a meta universe is applied to a user demand mining system, and comprises the following steps:
behavior habit mining is carried out on the metauniverse user activity big data, and a corresponding conversation behavior habit vector is obtained; the conversation behavior habit vector comprises user preference items of each activity event information in the metauniverse user activity big data;
determining visual windows corresponding to at least one user operation track in the metauniverse user activity big data respectively, and utilizing the determined at least one visual window to disassemble the conversation behavior habit vector to obtain stage behavior habit vectors corresponding to the at least one visual window respectively;
combining the user preference items of the activity event information and the first correlation scores among the user preference items to obtain corresponding service demand prediction knowledge;
combining the service demand forecasting knowledge and at least one stage behavior habit vector, and a second correlation score between the service demand forecasting knowledge and the at least one stage behavior habit vector to obtain a corresponding target session behavior habit vector and at least one target stage behavior habit vector;
and combining the target session behavior habit vector and the at least one target stage behavior habit vector to obtain a user demand theme corresponding to the metauniverse user activity big data.
In some independent embodiments, the combining the user preference items of the activity event information and the first correlation score between the user preference items to obtain the corresponding service demand prediction knowledge includes:
for each activity event message, the following steps are carried out:
taking one activity event information as main activity event information, and taking the rest activity event information in each activity event information as auxiliary activity event information;
based on a set AI rule, performing relevance score calculation on the user preference items of the main activity event information and the respective user preference items of the auxiliary activity event information respectively to obtain first relevance scores among the user preference items;
and integrating the user preference items of the activity event information according to the first relevance scores to obtain corresponding service demand prediction knowledge.
In some embodiments, the method further comprises, in combination with the service demand forecasting knowledge and at least one phase behavior habit vector, and a second correlation score between the service demand forecasting knowledge and the at least one phase behavior habit vector, obtaining a corresponding target session behavior habit vector and at least one target phase behavior habit vector, including:
for at least one phase behavior habit vector, the following steps are carried out:
based on a set AI rule, performing correlation scoring operation on user preference items of each activity event information in a stage behavior habit vector and activity demand knowledge of each activity event information in the service demand forecasting knowledge respectively to obtain a second correlation score between the stage behavior habit vector and the service demand forecasting knowledge;
and integrating the service demand forecasting knowledge and the at least one stage behavior habit vector respectively according to the second relevance scores to obtain a corresponding target conversation behavior habit vector and at least one target stage behavior habit vector.
In some independent embodiments, the obtaining, by combining the target conversation behavior habit vector and the at least one target phase behavior habit vector, the user requirement topic corresponding to the meta universe user activity big data includes:
respectively determining a scene requirement keyword corresponding to the target session behavior habit vector and a link requirement keyword corresponding to the at least one target stage behavior habit vector based on a requirement topic mining network;
and obtaining a user demand theme corresponding to the meta universe user activity big data according to the scene demand keyword and each link demand keyword.
In some embodiments that may be independent, the determining, based on the requirement topic mining network, a scene requirement keyword corresponding to the target session behavior habit vector and a link requirement keyword corresponding to the at least one target stage behavior habit vector, respectively, includes:
inputting the target conversation behavior habit vector into a requirement topic mining network, obtaining first hit indexes of the target conversation behavior habit vector, wherein the target conversation behavior habit vector belongs to each preset user requirement item, and taking each first hit index as a scene requirement keyword;
and respectively inputting the at least one target stage behavior habit vector into the requirement topic mining network, obtaining second hit indexes of the at least one target stage behavior habit vector, which respectively correspond to the preset user requirement items, and taking the second hit indexes as link requirement keywords.
In some embodiments that may be independent, the obtaining, according to the scene requirement keyword and each link requirement keyword, a user requirement theme corresponding to the meta universe user activity big data includes:
for each preset user requirement item, the following steps are implemented:
according to the second hit indexes of the at least one target stage behavior habit vector, which belong to a preset user requirement item, averaging the first hit indexes corresponding to the preset user requirement item and the second hit indexes meeting the preset judgment requirement, and determining the target hit index corresponding to the preset user requirement item;
and according to the target hit indexes corresponding to the preset user demand items, taking the preset user demand items with the target hit indexes larger than a preset limit value as user demand themes corresponding to the metauniverse user activity big data.
In some embodiments, the method further comprises the step of:
acquiring a debugging sample cluster; the debugging sample cluster comprises a plurality of user activity big data examples, and the user activity big data examples carry a priori requirement subjects;
circularly debugging the demand topic mining network by combining the debugging sample cluster until the preset debugging indexes are met;
wherein, the one-time cycle debugging process comprises: obtaining a target session behavior habit vector example and at least one target stage behavior habit vector example corresponding to the user activity big data example based on the user activity big data example screened from the debugging sample cluster;
determining a target requirement topic example corresponding to the user activity big data example according to the target conversation behavior habit vector example and the at least one target stage behavior habit vector example through the requirement topic mining network, and performing variable improvement on the requirement topic mining network by combining the target requirement topic example and the network cost determined by the prior requirement topic.
In some embodiments, the obtaining a target session behavior habit vector example and at least one target stage behavior habit vector example corresponding to the user activity big data example includes:
performing behavior habit mining on the user activity big data example to obtain a corresponding conversation behavior habit vector example, and based on at least one visual window obtained by determining at least one user operation trajectory example in the user activity big data example, disassembling the conversation behavior habit vector example to obtain stage behavior habit vector examples respectively corresponding to the at least one visual window example;
and obtaining a corresponding target conversation behavior habit vector example and at least one target stage behavior habit vector example according to the conversation behavior habit vector example and the at least one stage behavior habit vector example.
In some embodiments that may be independent, the determining, by the requirement topic mining network, a target requirement topic example corresponding to the user activity big data example according to the target session behavior habit vector example and the at least one target stage behavior habit vector example includes:
determining, by the requirement topic mining network, that the target session behavior habit vector examples respectively belong to first hit index examples of each user requirement item example, and the at least one target stage behavior habit vector example respectively corresponds to second hit index examples of each user requirement item example;
and determining a target requirement theme example corresponding to the user activity big data example according to the first hit index example and the second hit index example corresponding to each user requirement item example.
A user demand mining 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.
The method and the system for mining the user demand based on the metauniverse are used for mining behavior habits of metauniverse user activity big data to obtain corresponding conversation behavior habit vectors, resolving the conversation behavior habit vectors to obtain stage behavior habit vectors corresponding to the at least one visualization window respectively based on at least one visual window obtained by determining at least one user operation track in the metauniverse user activity big data, obtaining corresponding service demand prediction knowledge based on the service demand prediction knowledge and the at least one stage behavior habit vector and a second correlation score between the service demand prediction knowledge and the at least one stage behavior habit vector to obtain corresponding target conversation behavior habit vectors and at least one target stage behavior habit vector, and obtaining the user demand theme corresponding to the metauniverse user activity big data according to the target conversation behavior habit vectors and the at least one target stage behavior habit vector. In view of the fact that after the conversation behavior habit vector and the phase behavior habit vector of the Meta universe user activity big data are obtained, the conversation scene and the conversation link of the Meta universe user activity big data are in contact through demand prediction of the conversation behavior habit vector and the phase behavior habit vector, and therefore when the requirement theme is mined for the Meta universe user activity big data based on the obtained target conversation behavior habit vector and the target phase behavior habit vector, the user requirement theme which is accurate and credible as far as possible can be obtained, and the quality of user demand mining for the Meta universe user activity big data is improved.
Drawings
FIG. 1 is a schematic diagram illustrating one communication configuration of a customer demand mining system in which embodiments of the present application may be implemented.
FIG. 2 is a flow diagram illustrating a metastic-based user demand mining method that may implement an embodiment of the present application.
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 purpose, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the accompanying 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 making creative efforts 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 user demand mining system 100 that may implement embodiments of the present application, the user demand mining system 100 including a memory 101 for storing an executable computer program, and a processor 102 for implementing a metastic-based user demand mining method 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 requirement mining method based on a meta universe, which may implement an embodiment of the present application, where the user requirement mining method based on the meta universe may be implemented by the user requirement mining system 100 illustrated in fig. 1, and further may include the technical solutions described in the following steps 1 to 4.
Step 1, performing behavior habit mining on the metauniverse user activity big data to obtain a corresponding conversation behavior habit vector.
Wherein, the conversation behavior habit vector (which can be understood as behavior habit features in a global scenario) contains user preferences (which can be understood as behavior preferences of a user in a process of metaspace interaction) of each activity event information (such as a user behavior event) in the metaspace user activity big data (such as a user interaction activity record in a virtual mall). Furthermore, behavioral habit mining can be understood as feature extraction.
And 2, determining visual windows corresponding to at least one user operation track in the metauniverse user activity big data respectively, and utilizing the determined at least one visual window to disassemble the conversation behavior habit vector to obtain stage behavior habit vectors corresponding to the at least one visual window respectively.
For example, the user operation trajectory may be used to record some columns or consecutive operation behavior data of the user, and the corresponding visualization window may highlight the user operation trajectory, so that the session behavior habit vector may be split to obtain the local stage behavior habit vector.
And 3, combining the user preference items of the activity event information and the first correlation scores among the user preference items to obtain corresponding service demand prediction knowledge.
For example, a relevance score may be understood as a degree of association or a degree of relevance, and a relevance score may be a pearson relevance coefficient. Further, the service demand forecasting knowledge can focus on transition and forecasting from the behavior preference dimension to the user demand dimension, and the obtained service demand forecasting knowledge can reflect feature knowledge (feature vector) of the attention level of the user demand from the overall level.
And 4, combining the service demand forecasting knowledge and at least one stage behavior habit vector and a second correlation score between the service demand forecasting knowledge and the at least one stage behavior habit vector to obtain a corresponding target conversation behavior habit vector and at least one target stage behavior habit vector.
The target session behavior habit vector and the target stage behavior habit vector can reflect the characteristic vector of the behavior habit hiding the user requirement from the global level and the local level respectively, and therefore the target session behavior habit vector and the target stage behavior habit vector can be used as raw materials for user requirement theme mining.
And 5, combining the target conversation behavior habit vector and the at least one target stage behavior habit vector to obtain a user demand theme corresponding to the metauniverse user activity big data.
In the embodiment of the invention, the user requirement theme can be the category of the user requirement corresponding to the metauniverse user activity big data, such as 'identity privacy security', 'virtual commodity pushing', 'scene rendering optimization', and the like.
It can be understood that, when the method is applied to the steps 1 to 5, the conversation scene and the conversation link of the metauniverse user activity big data are associated by predicting the requirements of the conversation behavior habit vector and the stage behavior habit vector after the conversation behavior habit vector and the stage behavior habit vector of the metauniverse user activity big data are obtained, so that the user requirement theme which is as accurate and credible as possible can be obtained when the metauniverse user activity big data are subjected to requirement theme mining based on the obtained target conversation behavior habit vector and the target stage behavior habit vector, and the quality of the user requirement mining on the metauniverse user activity big data is improved.
Under some possible design considerations, the step 3 of obtaining the corresponding service demand forecasting knowledge by combining the user preference items of the activity event information and the first correlation scores among the user preference items may include the following steps: for each activity event message, the following steps are carried out: taking one activity event information as main activity event information (query event), and taking the rest of the activity event information in the activity event information as auxiliary activity event information (auxiliary event); based on a set AI rule (for example, a self-attention rule), performing correlation score operation on the user preference item of the main activity event information and the user preference item of each auxiliary activity event information (for example, the correlation score operation may be performed based on a Pearson correlation coefficient), to obtain a first correlation score (a value range may be set between 0 to 1) between the user preference items; and integrating (may be a weighted sum) the user preference items of the activity event information according to the first relevance scores to obtain corresponding service demand prediction knowledge. By the design, the corresponding service demand prediction knowledge can be accurately and completely obtained.
Under other possible design considerations, the step 4 of obtaining the corresponding target session behavior habit vector and at least one target phase behavior habit vector by combining the service demand forecasting knowledge and at least one phase behavior habit vector and the second correlation score between the service demand forecasting knowledge and the at least one phase behavior habit vector may include the following steps: for at least one phase behavior habit vector, the following steps are carried out: based on a set AI rule, performing correlation scoring operation on user preference items of each activity event information in a stage behavior habit vector and activity demand knowledge of each activity event information in the service demand forecasting knowledge respectively to obtain a second correlation score between the stage behavior habit vector and the service demand forecasting knowledge; and integrating the service demand forecasting knowledge and the at least one stage behavior habit vector respectively according to the second relevance scores to obtain a corresponding target conversation behavior habit vector and at least one target stage behavior habit vector. Therefore, the target conversation behavior habit vector and the target stage behavior habit vector which carry hidden demand details can be focused on based on the attention rules, and a credible basis is provided for subsequent user demand topic mining.
In some possible embodiments, the obtaining, by combining the target session behavior habit vector and the at least one target phase behavior habit vector, the user requirement theme corresponding to the metauniverse user activity big data, which is described in step 5, may include the technical solutions described in step 51 and step 52.
And step 51, based on the requirement topic mining network, respectively determining a scene requirement keyword corresponding to the target session behavior habit vector and a link requirement keyword corresponding to the at least one target stage behavior habit vector.
The requirement topic mining network can be a multi-classification network, such as a multiple regression model, the scene requirement keywords can be understood as global requirement labels, and the link requirement keywords can be understood as local requirement labels.
And step 52, obtaining a user requirement theme corresponding to the meta universe user activity big data according to the scene requirement keyword and each link requirement keyword.
It will be appreciated that, as applied to steps 51 and 52, different ranges of requirement keywords can be considered, thereby ensuring the suitability of the resulting user requirement topic.
In some possible embodiments, the determining, by the requirement topic mining network described in step 51, a scene requirement keyword corresponding to the target session behavior habit vector and a link requirement keyword corresponding to the at least one target stage behavior habit vector respectively may include the technical solutions described in step 511 and step 512.
Step 511, inputting the target conversation behavior habit vector into a requirement topic mining network, obtaining first hit indexes of the target conversation behavior habit vector, wherein the target conversation behavior habit vector belongs to each preset user requirement item, and taking each first hit index as a scene requirement keyword.
And step 512, inputting the at least one target stage behavior habit vector into the requirement topic mining network respectively, obtaining second hit indexes of the at least one target stage behavior habit vector corresponding to each preset user requirement item respectively, and taking the second hit indexes as link requirement keywords.
By way of example, the hit index may be understood as a probability value or a confidence score. In this way, based on steps 511 and 512, the link requirement keyword can be determined based on the hit index, thereby improving the accuracy and efficiency of determining the link requirement keyword.
In some examples, the obtaining of the user requirement theme corresponding to the meta universe user activity big data according to the scene requirement keyword and the link requirement keywords, which is described in step 52, may include the following: for each preset user requirement item, the following steps are implemented: according to the second hit indexes of the at least one target stage behavior habit vector, which belong to a preset user requirement item, averaging the first hit indexes corresponding to the preset user requirement item and the second hit indexes meeting the preset judgment requirement, and determining the target hit index corresponding to the preset user requirement item; and according to the target hit indexes corresponding to the preset user requirement items, taking the preset user requirement items with the target hit indexes larger than a preset limit value as user requirement themes corresponding to the metauniverse user activity big data. By the design, the precision of the determined user demand theme can be guaranteed by carrying out quantitative judgment based on the hit index, and the scene adaptability of the user demand theme can be guaranteed by carrying out analysis based on the preset user demand item.
Under some independent design ideas, the debugging process of the demand topic mining network can comprise the technical schemes described in S100 and S200.
S100, obtaining a debugging sample cluster.
The debugging sample cluster (debugging sample set) comprises a plurality of user activity big data examples (user activity big data samples), and the user activity big data examples carry a priori requirement theme (requirement theme truth value);
and S200, combining the debugging sample cluster, and circularly debugging the demand topic mining network until the demand topic mining network accords with a preset debugging index.
Wherein, the one-time cycle debugging process comprises: obtaining a target session behavior habit vector example and at least one target stage behavior habit vector example corresponding to the user activity big data example based on the user activity big data example screened from the debugging sample cluster; and determining a target demand topic example corresponding to the user activity big data example according to the target session behavior habit vector example and the at least one target stage behavior habit vector example through the demand topic mining network, and performing variable improvement on the demand topic mining network by combining the target demand topic example and the network cost determined by the prior demand topic.
The target conversation behavior habit vector example can be understood as a target conversation behavior habit vector sample, and the target stage behavior habit vector example can be understood as a target stage behavior habit vector sample. Other examples of information/data may be understood as training samples.
In other possible embodiments, obtaining a target session behavior habit vector example and at least one target phase behavior habit vector example corresponding to the user activity big data example includes: performing behavior habit mining on the user activity big data example to obtain a corresponding conversation behavior habit vector example, and based on at least one visual window obtained by determining at least one user operation trajectory example in the user activity big data example, disassembling the conversation behavior habit vector example to obtain stage behavior habit vector examples respectively corresponding to the at least one visual window example; and obtaining a corresponding target conversation behavior habit vector example and at least one target stage behavior habit vector example according to the conversation behavior habit vector example and the at least one stage behavior habit vector example.
On the basis of the above, the determining, by the requirement topic mining network, the target requirement topic example corresponding to the user activity big data example according to the target session behavior habit vector example and the at least one target stage behavior habit vector example includes: determining, by the requirement topic mining network, that the target session behavior habit vector examples respectively belong to first hit index examples of each user requirement item example, and the at least one target stage behavior habit vector example respectively corresponds to second hit index examples of each user requirement item example; and determining a target requirement subject example corresponding to the user activity big data example according to the first hit index example and the second hit index example corresponding to each user requirement item example.
By the aid of the debugging thought, network training can be performed based on network cost, and accordingly precision of the demand topic mining network in the operation process is guaranteed.
The embodiments of the present application have been described above with reference to the accompanying drawings, and have at least the following beneficial effects: the method and the system for mining the user demand based on the metauniverse are used for mining behavior habits of metauniverse user activity big data to obtain corresponding conversation behavior habit vectors, resolving the conversation behavior habit vectors to obtain stage behavior habit vectors corresponding to the at least one visualization window respectively based on at least one visual window obtained by determining at least one user operation track in the metauniverse user activity big data, obtaining corresponding service demand prediction knowledge based on the service demand prediction knowledge and the at least one stage behavior habit vector and a second correlation score between the service demand prediction knowledge and the at least one stage behavior habit vector to obtain corresponding target conversation behavior habit vectors and at least one target stage behavior habit vector, and obtaining the user demand theme corresponding to the metauniverse user activity big data according to the target conversation behavior habit vectors and the at least one target stage behavior habit vector. In view of the fact that after the conversation behavior habit vector and the phase behavior habit vector of the Meta universe user activity big data are obtained, the conversation scene and the conversation link of the Meta universe user activity big data are in contact through demand prediction of the conversation behavior habit vector and the phase behavior habit vector, and therefore when the requirement theme is mined for the Meta universe user activity big data based on the obtained target conversation behavior habit vector and the target phase behavior habit vector, the user requirement theme which is accurate and credible as far as possible can be obtained, and the quality of user demand mining for the Meta universe user activity big data is improved.
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 (10)
1. A user demand mining method based on a meta universe is characterized by being applied to a user demand mining system and comprising the following steps:
behavior habit mining is carried out on the meta universe user activity big data, and a corresponding conversation behavior habit vector is obtained; the conversation behavior habit vector contains user preference items of each activity event information in the meta universe user activity big data;
determining visual windows corresponding to at least one user operation track in the metauniverse user activity big data respectively, and utilizing the determined at least one visual window to disassemble the conversation behavior habit vector to obtain stage behavior habit vectors corresponding to the at least one visual window respectively;
combining the user preference items of the activity event information and the first correlation scores among the user preference items to obtain corresponding service demand prediction knowledge;
combining the service demand forecasting knowledge and at least one stage behavior habit vector, and a second correlation score between the service demand forecasting knowledge and the at least one stage behavior habit vector to obtain a corresponding target session behavior habit vector and at least one target stage behavior habit vector;
and combining the target session behavior habit vector and the at least one target stage behavior habit vector to obtain a user demand theme corresponding to the metauniverse user activity big data.
2. The method of claim 1, wherein the obtaining of the corresponding service demand forecasting knowledge in combination with the user preferences of the activity event information and the first correlation score between the user preferences comprises:
for each activity event message, the following steps are carried out:
taking one activity event information as main activity event information, and taking the rest of the activity event information in each activity event information as auxiliary activity event information;
based on a set AI rule, performing relevance score calculation on the user preference items of the main activity event information and the respective user preference items of the auxiliary activity event information respectively to obtain a first relevance score among the user preference items;
and integrating the user preference items of the activity event information according to the first relevance scores to obtain corresponding service demand prediction knowledge.
3. The method of claim 1, wherein combining the service demand forecasting knowledge and at least one stage behavior habit vector, and a second correlation score between the service demand forecasting knowledge and the at least one stage behavior habit vector to obtain a corresponding target conversational behavior habit vector and at least one target stage behavior habit vector comprises:
for at least one phase behavior habit vector, the following steps are carried out:
based on a set AI rule, performing correlation scoring operation on user preference items of each activity event information in a stage behavior habit vector and activity demand knowledge of each activity event information in the service demand forecasting knowledge respectively to obtain a second correlation score between the stage behavior habit vector and the service demand forecasting knowledge;
and integrating the service demand forecasting knowledge and the at least one stage behavior habit vector respectively according to the second relevance scores to obtain a corresponding target conversation behavior habit vector and at least one target stage behavior habit vector.
4. The method according to claim 1, wherein the obtaining the user demand topic corresponding to the metauniverse user activity big data by combining the target session behavior habit vector and the at least one target phase behavior habit vector comprises:
based on a requirement topic mining network, respectively determining a scene requirement keyword corresponding to the target session behavior habit vector and a link requirement keyword corresponding to the at least one target stage behavior habit vector;
and obtaining a user demand theme corresponding to the metauniverse user activity big data according to the scene demand keyword and each link demand keyword.
5. The method of claim 4, wherein the determining, based on the requirement topic mining network, the scenario requirement keyword corresponding to the target session behavior habit vector and the link requirement keyword corresponding to the at least one target stage behavior habit vector respectively comprises:
inputting the target conversation behavior habit vector into a requirement topic mining network, obtaining first hit indexes of the target conversation behavior habit vector, wherein the target conversation behavior habit vector belongs to each preset user requirement item, and taking each first hit index as a scene requirement keyword;
and respectively inputting the at least one target stage behavior habit vector into the requirement topic mining network, obtaining second hit indexes of the at least one target stage behavior habit vector, which respectively correspond to the preset user requirement items, and taking the second hit indexes as link requirement keywords.
6. The method according to claim 5, wherein the obtaining of the user requirement theme corresponding to the metauniverse user activity big data according to the scene requirement keyword and each link requirement keyword comprises:
for each preset user requirement item, the following steps are implemented:
averaging the second hit indexes meeting the preset judgment requirement and the first hit indexes corresponding to the preset user requirement items according to the second hit indexes of the behavior habit vectors of the at least one target stage, which belong to one preset user requirement item, and determining the target hit indexes corresponding to the preset user requirement items;
and according to the target hit indexes corresponding to the preset user requirement items, taking the preset user requirement items with the target hit indexes larger than a preset limit value as user requirement themes corresponding to the metauniverse user activity big data.
7. The method of claim 4, wherein the commissioning process of the demand topic mining network comprises:
acquiring a debugging sample cluster; the debugging sample cluster comprises a plurality of user activity big data examples, and the user activity big data examples carry a priori requirement subjects;
circularly debugging the demand topic mining network by combining the debugging sample cluster until the preset debugging indexes are met;
wherein, the one-time cycle debugging process comprises: obtaining a target session behavior habit vector example and at least one target stage behavior habit vector example corresponding to the user activity big data example based on the user activity big data example screened from the debugging sample cluster;
determining a target requirement topic example corresponding to the user activity big data example according to the target conversation behavior habit vector example and the at least one target stage behavior habit vector example through the requirement topic mining network, and performing variable improvement on the requirement topic mining network by combining the target requirement topic example and the network cost determined by the prior requirement topic.
8. The method according to claim 7, wherein the obtaining of the target conversation behavior habit vector example and the at least one target stage behavior habit vector example corresponding to the user activity big data example comprises:
performing behavior habit mining on the user activity big data example to obtain a corresponding conversation behavior habit vector example, and based on at least one visual window obtained by determining at least one user operation trajectory example in the user activity big data example, disassembling the conversation behavior habit vector example to obtain stage behavior habit vector examples respectively corresponding to the at least one visual window example;
and obtaining a corresponding target conversation behavior habit vector example and at least one target stage behavior habit vector example according to the conversation behavior habit vector example and the at least one stage behavior habit vector example.
9. The method according to claim 7, wherein the determining, by the requirement topic mining network, a target requirement topic example corresponding to the user activity big data example according to the target conversation behavior habit vector example and the at least one target stage behavior habit vector example comprises:
determining, by the requirement topic mining network, that the target session behavior habit vector examples respectively belong to first hit index examples of each user requirement item example, and the at least one target stage behavior habit vector example respectively corresponds to second hit index examples of each user requirement item example;
and determining a target requirement subject example corresponding to the user activity big data example according to the first hit index example and the second hit index example corresponding to each user requirement item example.
10. A user demand mining 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.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211283432.9A CN115712657A (en) | 2022-10-20 | 2022-10-20 | User demand mining method and system based on meta universe |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211283432.9A CN115712657A (en) | 2022-10-20 | 2022-10-20 | User demand mining method and system based on meta universe |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115712657A true CN115712657A (en) | 2023-02-24 |
Family
ID=85229956
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211283432.9A Withdrawn CN115712657A (en) | 2022-10-20 | 2022-10-20 | User demand mining method and system based on meta universe |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115712657A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116307218A (en) * | 2023-03-27 | 2023-06-23 | 松原市邹佳网络科技有限公司 | Meta-universe experience user behavior prediction method and system based on artificial intelligence |
| CN116958447A (en) * | 2023-08-09 | 2023-10-27 | 深圳市固有色数码技术有限公司 | Automatic meta-universe character generation system and method based on Internet of things |
| CN117149986A (en) * | 2023-10-31 | 2023-12-01 | 杭州海兴泽科信息技术有限公司 | Real-time big data processing method and system based on multi-stage data channel |
-
2022
- 2022-10-20 CN CN202211283432.9A patent/CN115712657A/en not_active Withdrawn
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116307218A (en) * | 2023-03-27 | 2023-06-23 | 松原市邹佳网络科技有限公司 | Meta-universe experience user behavior prediction method and system based on artificial intelligence |
| CN116958447A (en) * | 2023-08-09 | 2023-10-27 | 深圳市固有色数码技术有限公司 | Automatic meta-universe character generation system and method based on Internet of things |
| CN117149986A (en) * | 2023-10-31 | 2023-12-01 | 杭州海兴泽科信息技术有限公司 | Real-time big data processing method and system based on multi-stage data channel |
| CN117149986B (en) * | 2023-10-31 | 2024-02-09 | 杭州海兴泽科信息技术有限公司 | Real-time big data processing method and system based on multi-stage data channel |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10958748B2 (en) | Resource push method and apparatus | |
| US12099558B2 (en) | Method and apparatus for providing content based on user activity | |
| CN107463704B (en) | Search method and device based on artificial intelligence | |
| CN111782965A (en) | Intention recommendation method, device, equipment and storage medium | |
| CN115712657A (en) | User demand mining method and system based on meta universe | |
| CN110019616B (en) | POI (Point of interest) situation acquisition method and equipment, storage medium and server thereof | |
| CN109657054A (en) | Abstraction generating method, device, server and storage medium | |
| CN111597446B (en) | Content pushing method and device based on artificial intelligence, server and storage medium | |
| CN110598109A (en) | Information recommendation method, device, equipment and storage medium | |
| US20210232972A1 (en) | Methods, mediums, and systems for an unsupervised predictive learning system | |
| CN115062732A (en) | Resource sharing cooperation recommendation method and system based on big data user tag information | |
| CN114691828A (en) | Data processing method, device, equipment and medium | |
| CN110598126B (en) | Cross-social network user identity recognition method based on behavior habits | |
| CN114242047B (en) | Voice processing method and device, electronic equipment and storage medium | |
| CN116484105B (en) | Service processing method, device, computer equipment, storage medium and program product | |
| KR20210009885A (en) | Method, device and computer readable storage medium for automatically generating content regarding offline object | |
| US20230252980A1 (en) | Multi-channel conversation processing | |
| CN114638308A (en) | Method and device for acquiring object relationship, electronic equipment and storage medium | |
| CN115017362A (en) | Data processing method, electronic device and storage medium | |
| CN112883256B (en) | Multitasking method, apparatus, electronic device and storage medium | |
| CN111078972B (en) | Questioning behavior data acquisition method, questioning behavior data acquisition device and server | |
| CN113505293A (en) | Information pushing method and device, electronic equipment and storage medium | |
| CN114417875B (en) | Data processing method, apparatus, device, readable storage medium, and program product | |
| CN113343142B (en) | News click rate prediction method based on user behavior sequence filling and screening | |
| CN116541601B (en) | Method, device, equipment and medium for improving near line recall capability of recommendation system |
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: 20230224 |