CN116910362B - Intelligent recommendation method for perceived data, computer equipment and storage medium - Google Patents
Intelligent recommendation method for perceived data, computer equipment and storage medium Download PDFInfo
- Publication number
- CN116910362B CN116910362B CN202310876923.2A CN202310876923A CN116910362B CN 116910362 B CN116910362 B CN 116910362B CN 202310876923 A CN202310876923 A CN 202310876923A CN 116910362 B CN116910362 B CN 116910362B
- Authority
- CN
- China
- Prior art keywords
- target user
- data
- information
- interest
- recommendation
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000003860 storage Methods 0.000 title claims abstract description 13
- 238000012544 monitoring process Methods 0.000 claims abstract description 92
- 238000004458 analytical method Methods 0.000 claims abstract description 90
- 230000008447 perception Effects 0.000 claims abstract description 42
- 238000012545 processing Methods 0.000 claims abstract description 17
- 230000003993 interaction Effects 0.000 claims abstract description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 33
- 238000004364 calculation method Methods 0.000 claims description 26
- 238000009826 distribution Methods 0.000 claims description 26
- 230000006399 behavior Effects 0.000 claims description 16
- 238000011156 evaluation Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 11
- 241000283153 Cetacea Species 0.000 claims description 10
- 238000000513 principal component analysis Methods 0.000 claims description 10
- 238000001914 filtration Methods 0.000 claims description 9
- 238000012821 model calculation Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 5
- 239000000047 product Substances 0.000 abstract description 44
- 238000007405 data analysis Methods 0.000 abstract description 8
- 239000012467 final product Substances 0.000 abstract description 4
- 230000005540 biological transmission Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000011157 data evaluation Methods 0.000 description 1
- 238000013079 data visualisation Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a perception data intelligent recommendation method, computer equipment and a storage medium, a target user information monitoring platform is established, a target user personalized feature data set is acquired through an information acquisition knowledge graph establishing end, the target user personalized feature data set is transmitted to an information perception predicting end for semantic analysis and label establishment, accurate target user personalized feature monitoring information is obtained, data analysis is carried out through a target user interest monitoring analysis model, target user interest recommendation data is obtained, when the interest recommendation data is smaller than or equal to an interest recommendation coefficient node value, an interest adjustment signal is sent, and a self-adaptive intelligent recommending end is utilized for human-computer interaction adjustment management, so that the technical problems that the monitoring analysis method for a target user is not intelligent enough, the data analysis flow and the processing dimension are insufficient in the prior art, the final product monitoring recommendation result is not accurate are solved, intelligent monitoring analysis of a target user product of a recommendation system is realized, and recommendation accuracy is ensured.
Description
Technical Field
The present invention relates to the field of recommendation system applications, and in particular, to a method for intelligently recommending perceptual data, a computer device, and a storage medium.
Background
The recommendation system is an information retrieval tool in the Internet era, people recognize the value of the recommendation system since the 90 th century, and after more than twenty years of accumulation and precipitation, the recommendation system gradually becomes an independent subject and achieves a lot of results in academic research and industrial application.
The recommendation system can be regarded as a multi-classification problem of machine learning, how to find out products like by customers from various products becomes a main target of the recommendation system algorithm design, and early recommendation systems mainly apply collaborative filtering algorithms, which do not need to obtain characteristic data of users or articles in advance, and model the users only depending on historical behavior data of the users so as to recommend the users. Collaborative filtering algorithms mainly include user-based collaborative filtering, item-based collaborative filtering, a lingo-semantic model, and the like.
The target user is used as a key of the application of the recommendation system and is an essential component for maintaining the normal operation of the recommendation system, and at present, the target user personalized monitoring is carried out by mainly carrying out integrated management and control on the target user and monitoring the target user personalized characteristics in real time, but the interest monitoring adjustment result of the target user cannot reach the expected standard due to the limitation of the prior art.
In the prior art, when the interest of the target user is monitored and adjusted, the monitoring and analysis method is not intelligent enough, and the data analysis flow and the processing dimension are not enough, so that the final product monitoring and recommending result is not accurate enough.
Disclosure of Invention
The application provides an intelligent recommendation method for perception data, computer equipment and a storage medium, which are used for solving the technical problem that the monitoring and analyzing method is not intelligent enough and the data analysis flow and the processing dimension are insufficient when the interest of a target user is monitored and adjusted in the prior art, so that the final product monitoring and recommending result is not accurate enough.
In view of the above problems, the application provides an intelligent recommendation method for perceived data, a computer device and a storage medium.
The application provides an intelligent recommendation method for perception data, which comprises the following steps: establishing a target user information monitoring platform, wherein the target user monitoring platform comprises an information acquisition knowledge graph establishing end, an information perception predicting end and a self-adaptive intelligent recommending end; acquiring a target user personalized feature data set through the information acquisition knowledge graph establishing end, wherein the target user personalized feature data set comprises target user unit time preference information Qe, target user unit time behavior information Fz and target user unit time character information Tr; the target user unit time preference information Qe, the target user unit time behavior information Fz and the target user unit time character information Tr are transmitted to the information perception prediction end after the information acquisition knowledge graph establishment end establishes a personal information knowledge graph through different factor weight models; semantic analysis and label establishment are carried out on the target user personalized feature data set through the information perception prediction end, so that accurate target user personalized feature monitoring information Jb= [ J1, J2., jn ] is obtained; obtaining a target user interest monitoring analysis model by utilizing the information perception prediction end, and inputting the accurate target user personalized feature monitoring information Jb= [ J1, J2, & gt, jn ] into the target user interest monitoring analysis model for analysis to obtain target user interest recommendation data; setting an interest recommendation coefficient node value, and sending an interest adjustment signal when the interest recommendation data of the target user is less than or equal to the interest recommendation coefficient node value; and carrying out man-machine interaction adjustment management on the interest recommendation data of the target user by utilizing the interest adjustment signal and the self-adaptive intelligent recommendation terminal.
The personal information knowledge graph is established through different factor weight models, and the expression is as follows:
Wherein represents a connection function between nodes of the knowledge graph, K represents a connection factor between the nodes, N represents a single node, N represents the total node number of the knowledge graph, P represents the connection path of the single node of the knowledge graph, represents the dynamic distribution of the connection paths of different single nodes, xi, ω and λ respectively represent the preference information of the unit time of the target user, the behavior information of the unit time of the target user and the weight coefficient of the character information of the unit time of the target user when the knowledge graph is constructed, and/ represents the connection error between the nodes;
The target user interest monitoring analysis model has the expression:
wherein represents a target user interest recommendation function, θ represents a feature factor of interest tags after semantic analysis, Y represents iteration times, Y represents a total number of target user interest categories,/> represents target user interest distribution, and/> represents a target user feature analysis error;
The application provides an intelligent recommendation method for perception data, which comprises the following steps: the target user monitoring platform building unit is used for building a target user information monitoring platform, and comprises an information acquisition knowledge graph building end, an information perception prediction end and a self-adaptive intelligent recommendation end; the target user personalized feature data set establishing unit is used for acquiring a target user personalized feature data set through the information acquisition knowledge graph establishing end, and the target user personalized feature data set comprises target user unit time preference information Qe, target user unit time behavior information Fz and target user unit time character information Tr; the target user personalized feature data transmission unit is used for establishing a personal information knowledge graph through the information acquisition knowledge graph establishment end and then sending the personal information knowledge graph to the information perception prediction end through different factor weight models; the semantic analysis and label establishment unit is used for carrying out semantic analysis and label establishment on the target user personalized feature data set through the information perception prediction end to obtain accurate target user personalized feature monitoring information Jb= [ J1, J2, the first place, jn ]; the user interest monitoring and analyzing unit is used for obtaining a target user interest monitoring and analyzing model by utilizing the information perception prediction end, inputting the accurate target user personalized feature monitoring information Jb= [ J1, J2, ], jn ] into the target user interest monitoring and analyzing model for analysis, and obtaining target user interest recommendation data; the node value judging unit is used for setting an interest recommendation coefficient node value, and sending an interest adjustment signal when the interest recommendation data of the target user is less than or equal to the interest recommendation coefficient node value; and the man-machine interaction adjustment management unit is used for carrying out man-machine interaction adjustment management on the interest recommendation data of the target user by utilizing the interest adjustment signal and the self-adaptive intelligent recommendation terminal.
The computer equipment comprises a memory and a processor, wherein the memory stores a computer program, and the computer equipment is characterized in that the processor realizes the intelligent recommendation method of the perception data when executing the computer program.
The invention relates to a computer readable storage medium storing a computer program, which is characterized in that the computer program realizes the intelligent recommendation method of the perception data when being executed by a processor.
The beneficial effects are that: the intelligent recommendation method for perceived data provided by the embodiment of the application establishes a target user information monitoring platform, acquires and obtains a target user personalized feature data set through an information acquisition knowledge graph establishing end, comprises target user unit time preference information, target user unit time behavior information and target user unit time character information, and sends the target user personalized feature data set to an information perception predicting end after establishing a personal information knowledge graph through the information acquisition knowledge graph establishing end through different factor weight models, performs semantic analysis and label establishment to obtain accurate target user personalized feature monitoring information, establishes a target user interest monitoring analysis model, obtains target user interest recommendation data through model analysis, sets interest recommendation coefficient node values, sends out interest adjustment signals when the target user interest recommendation data is smaller than or equal to the interest recommendation coefficient node values, and further performs man-machine interaction adjustment management on the target user interest recommendation data through a self-adaptive intelligent recommendation end, the accuracy of recommendation is ensured.
Drawings
FIG. 1 is a first flowchart of the method of the present invention;
FIG. 2 is a second flowchart of the method of the present invention;
FIG. 3 is a third flowchart of the method of the present invention;
FIG. 4 is a fourth flowchart of the method of the present invention;
FIG. 5 is a fifth flowchart of the method of the present invention;
fig. 6 is a block diagram of a method of the present invention.
Detailed Description
The application provides a perception data intelligent recommendation method, computer equipment and storage medium, a target user information monitoring platform is established, an information acquisition knowledge graph establishing end is used for acquiring a target user personalized characteristic data set, an information acquisition knowledge graph establishing end is used for establishing a personal information knowledge graph through different factor weight models and then sending the personal information knowledge graph to an information perception predicting end, semantic analysis and label establishment are carried out to obtain accurate target user personalized characteristic monitoring information Jb= [ J1, J2, & gt, jn ], a target user interest monitoring analysis model is established, target user interest recommendation data is obtained through model analysis, when the target user interest recommendation data is less than or equal to an interest recommendation coefficient node value, an interest adjustment signal is sent, and human-computer interaction adjustment management is carried out by utilizing an adaptive intelligent recommendation end, so that the problem that a final product recommendation result is not accurate due to insufficient intelligent monitoring analysis methods and processing dimensions in the prior art is solved.
As shown in fig. 1, the present application provides an intelligent recommendation method for sensing data, which includes:
Step A1: establishing a target user information monitoring platform, wherein the target user monitoring platform comprises an information acquisition knowledge graph establishing end, an information perception predicting end and a self-adaptive intelligent recommending end;
Specifically, the intelligent recommendation method for perceived data provided by the application is characterized in that the real-time personalized characteristic data set of the target user is acquired, the invalid data is reduced on the basis of guaranteeing the completeness of the data by data enclosing optimization, further, the node value of the data analysis and interest recommendation coefficient is judged, the target user with potential interest risk is regulated and warned, the node value of the interest recommendation coefficient is a coefficient which is established by recommending the target user according to interest classification, the node is related to different unit time, the node is a time node, firstly, the information monitoring platform of the target user is established, the information monitoring platform of the recommended scene is a multi-level platform, the information acquisition knowledge map establishing end, the information sensing prediction end and the self-adaptive intelligent recommendation end penetrate through the complete process of interest regulation analysis of the target user, the information acquisition knowledge map establishing end is used for carrying out data acquisition and transmission of the target user, the information sensing prediction end is used for carrying out data semantic analysis and label establishment, interest analysis is carried out on the processed data recommendation precision, the information is subjected to interest analysis, the self-adaptive intelligent prediction end is used for carrying out interest regulation and warning information is used for displaying the target user interest regulation and real-time, and the target user interest regulation is carried out on the target information regulation and the target information is displayed.
Step A2: acquiring a target user personalized feature data set through the information acquisition knowledge graph establishing end, wherein the target user personalized feature data set comprises target user unit time preference information Qe, target user unit time behavior information Fz and target user unit time character information Tr;
Step A3: the target user unit time preference information Qe, the target user unit time behavior information Fz and the target user unit time character information Tr are transmitted to the information perception prediction end after the information acquisition knowledge graph establishment end establishes a personal information knowledge graph through different factor weight models;
Specifically, the information collection knowledge graph establishing end is utilized to collect data of the target user unit time preference information Qe, the target user unit time behavior information Fz and the target user unit time character information Tr, wherein the target user unit time preference information Qe comprises target user movement information, entertainment information, adventure information, intelligence information, collection information, musical instruments information, literature information and the like; the target user unit time behavior information Fz comprises purchase times, purchase frequency, product browsing times and the like; the target user unit time personality information Tr comprises target optimistic, outward, open, satisfied, excited, homonymy, poor, self-confidence, responsibility, cautiousness, doubt and the like, and through collecting personalized characteristic data sets of a plurality of target users in a recommended scene, and carrying out corresponding identification of the target users and the collected personalized characteristic data sets so as to carry out subsequent identification and differentiation, the data information is integrated to generate the target user personalized characteristic data set, further, the information collection knowledge graph establishing end is utilized to carry out encryption transmission on the target user unit time preference information Qe, the target user unit time behavior information Fz and the target user unit time personality information Tr, and the information perception predicting end is utilized to receive the data, so that the data privacy can be effectively ensured, the information leakage is avoided, and the information perception predicting end is utilized to carry out data analysis processing.
Establishing a personal information knowledge graph through different factor weight models, wherein the expression is as follows:
Wherein represents a connection function between nodes of the knowledge graph, K represents a connection factor between the nodes, N represents a single node, N represents the total node number of the knowledge graph, P represents the connection path of the single node of the knowledge graph, represents the dynamic distribution of the connection paths of different single nodes, xi, ω and λ respectively represent the preference information of the unit time of the target user, the behavior information of the unit time of the target user and the weight coefficient of the character information of the unit time of the target user when the knowledge graph is constructed, and/ represents the connection error between the nodes;
Step A4: semantic analysis and label establishment are carried out on the target user personalized feature data set through the information perception prediction end, so that accurate target user personalized feature monitoring information Jb= [ J1, J2., jn ] is obtained;
Specifically, the information collection knowledge graph establishment end is used for transmitting the target user personalized feature data set to the information perception prediction end, data principal component analysis is carried out on the target user personalized feature data set, correlation data check is further carried out, whether the correlation data can be changed or not is judged, pearson correlation coefficient calculation is carried out on the correlation data which can be changed by using the correlation among the data, rejection processing is carried out on the correlation data which cannot be changed, optimization is carried out on the correlation data so as to improve data completeness, further, bat algorithm local positioning is carried out on the data after the pearson correlation coefficient calculation is carried out, the clarity and data regularity of data visualization are improved, and the target user personalized feature data set after semantic analysis and label establishment is used as accurate target user personalized feature monitoring information Jb= [ J1, J2, jn ].
As shown in fig. 2, performing semantic analysis and label establishment on the target user personalized feature data set by the information perception prediction end, to obtain accurate target user personalized feature monitoring information jb= [ J1, J2, and Jn ] further includes:
Setting a personalized feature processing algorithm, wherein the personalized feature processing algorithm comprises principal component analysis, whale algorithm tracking and bat algorithm local positioning;
Performing principal component analysis on the target user personalized feature data set to obtain target user personalized feature data information with correlation;
tracking a whale algorithm according to the target user personalized feature data information with correlation to obtain a target user personalized feature data information boundary range;
And carrying out bat algorithm local positioning on the target user personalized feature data information boundary range by utilizing a data fluctuation range to obtain the accurate target user personalized feature monitoring information Jb= [ J1, J2 ].
Specifically, the target user personalized feature data set is transmitted to the information perception prediction end, semantic analysis and label establishment are carried out on the target user personalized feature data set by utilizing the information perception prediction end, the principal component analysis, the whale algorithm tracking and the bat algorithm local positioning are used as data semantic analysis and label establishment processes, the data semantic analysis and label establishment processes are set as the personalized feature processing algorithm, the principal component analysis is carried out on the target user personalized feature data set, the data are in the same order of magnitude by data standardization, the subsequent comparison analysis is convenient, the target user personalized feature data information with correlation is obtained, the target user personalized feature data with correlation is further traversed, abnormal data in the target user personalized feature data are examined by utilizing search conditions, the accuracy of the subsequent data analysis and the actual agreement of the data are improved, for example, the dynamic change trend of the data is determined, the data which is free from the trend change main curve is removed, the whale algorithm tracking is carried out on the basis of not influencing the data information quantity, the personalized characteristic data information boundary range of the target user is obtained, further, the fluctuation range of the data, namely the data representation format, for example, the data type, the data format and the like, is determined, the bat algorithm local positioning is carried out on the personalized characteristic data information boundary range of the target user by taking the fluctuation range as a standard, the secondary standardization arrangement is carried out on the data, the neatness and the definition of the data visual display can be effectively improved, the personalized characteristic monitoring information Jb= [ J1, J2, ], jn of the accurate target user is obtained, the personalized characteristic monitoring information Jb= [ J1 ] of the accurate target user, and J2, jn is used as a data semantic analysis and label establishment result, so that convenience is provided for subsequent data identification analysis processing.
The target user interest monitoring analysis model has the expression:
Wherein represents a target user interest recommendation function, θ represents a feature factor of interest tags after semantic analysis, Y represents iteration times, Y represents a total number of target user interest categories,/> represents target user interest distribution, and/> represents a target user feature analysis error;
as shown in fig. 3, further, performing whale algorithm tracking according to the target user personalized feature data information with correlation, and obtaining the boundary range of the target user personalized feature data information further includes:
acquiring data search conditions, and carrying out data surrounding on the target user personalized feature data information with correlation by utilizing the data search conditions to acquire effective target user personalized feature data information;
coordination and filtration are carried out on the personalized characteristic data information of the effective target user, and the user similarity of the recommended data is obtained;
When the recommendation data user similarity reaches the preset data user similarity, carrying out pearson correlation coefficient calculation on the effective target user personalized feature data information to obtain correction target user personalized feature data information;
And carrying out feature coding on the personalized feature data information of the corrected target user to obtain the boundary range of the personalized feature data information of the target user.
Specifically, main component analysis is performed on the original acquired data to obtain the target user personalized feature data information with correlation, the data search condition is established, the data search condition is a screening condition standard of data, such as abnormal data, repeated data, irrelevant data and the like, the data search condition is used for cleaning the target user personalized feature data information with correlation, integration processing is performed on residual data after the investigation to obtain the effective target user personalized feature data information, whether the effective target user personalized feature data has a data fault or not is further judged, proportion calculation is performed on the correlation data and the total data, and the user similarity of the recommended data is determined through data integrity evaluation.
Further, the preset data user similarity is obtained, the preset data user similarity is a user similarity critical value for judging the size of the data missing amount, for example, ten percent can be set to be the preset data user similarity, further whether the recommended data user similarity reaches the preset data user similarity is judged, when the recommended data user similarity reaches the preset data user similarity, the data user similarity is excessively high, the influence on the subsequent data analysis and evaluation is caused, the correlation data is optimized to improve the data information amount, the personalized characteristic data information of the corrected target user is obtained, the position of the personalized characteristic data information of the corrected target user is further matched with the position of the correlation data, the data specification integration processing is performed after the deletion pearson correlation coefficient is calculated, the boundary range of the personalized characteristic data information of the target user is obtained, invalid data can be removed on the basis of guaranteeing the completeness of the data information through data filtering and missing optimization, and the data integrity is improved.
As shown in fig. 4, further, when the recommendation data user similarity reaches a preset data user similarity, performing pearson correlation coefficient calculation on the effective target user personalized feature data information to obtain correction target user personalized feature data information further includes:
Carrying out item average score calculation on recommendation correlation data information of the personalized feature data information of the effective target user to obtain item average scores;
when the average score of the article is larger than a preset average score, obtaining a correlation data score bias;
Calculating the pearson correlation coefficient of the recommended correlation data information by utilizing a data co-occurrence matrix and the correlation data scoring bias; and when the average score of the article is less than or equal to the preset average score, carrying out data re-acquisition optimization on the recommended correlation data information.
Specifically, the data user similarity evaluation is carried out on the personalized feature data information of the effective target user, the recommended correlation data information is obtained, whether a modifiable space exists in the recommended correlation data or not is further judged, namely, the necessity of original pearson correlation coefficient calculation is determined, the influence degree of optimizing the data information by using similar data is determined, the article average score is obtained through carrying out alterable evaluation analysis, further, the preset average score is obtained, namely, the critical value of the article average score is limited, when the article average score is larger than the preset average score, the correlation data is indicated to be properly changed and is in a controllable interval, the correlation data score bias is collected, namely, the correlation data is distributed in the personalized feature data information of the effective target user, adjacent data information and the like, a data co-occurrence matrix is determined, namely, the correlation relationship between the correlation data and the adjacent data and the total data amount is determined, for example, the data type, the data change trend, the data change scale standard and the like, the correlation data information is optimized through carrying out comprehensive correlation analysis, the information is improved, when the preset average score is larger than the preset average score, the correlation data can be collected in a controllable interval, the random influence on the correlation coefficient calculation is not carried out on the pearson correlation coefficient calculation by carrying out calculation, the random influence on the correlation data is carried out on the correlation coefficient calculation by using the binary analysis, and the optimization efficiency is improved.
As shown in fig. 5, when the recommendation data user similarity reaches a preset data user similarity, performing pearson correlation coefficient calculation on the effective target user personalized feature data information, and obtaining correction target user personalized feature data information further includes:
Searching a preset distribution range of the personalized characteristic data information of the effective target user by utilizing the correlation data grading bias to obtain distribution correlation data information;
Carrying out correlation evaluation on the distribution correlation data information to obtain a correlation coefficient of the correlation data;
Establishing a correlation data ordering model according to the correlation coefficient of the correlation data and the distribution correlation data information;
and calculating the pearson correlation coefficient of the recommended correlation data information by using the correlation data ordering model.
Specifically, the distribution position of the correlation data is subjected to positioning identification, the correlation data scoring bias is obtained, the preset distribution area is obtained, namely, a data interval intercepted by the correlation data information is obtained, the range retrieval is carried out on the personalized characteristic data of the effective target user by utilizing the preset distribution area, a plurality of correlation data distribution intervals including the correlation data and adjacent correlation data are obtained, the correlation data information is further subjected to correlation evaluation, the correlation degree between the correlation data in the missing distribution data information and other multiple correlation data in the distribution area is determined, the correlation data correlation coefficient is obtained, the data correlation degree is in direct proportion to the correlation coefficient, the correlation data correlation coefficient and the distribution correlation data information are further utilized to establish a correlation data sequencing model, an analysis function is arranged in the correlation data sequencing model, the correlation data can be directly generated by utilizing input information, the correlation data is used as alternative data to carry out pearson correlation coefficient calculation on the recommended correlation data information, the correlation coefficient calculation is carried out by calculating the correlation coefficient of the correlation data, the correlation coefficient of the correlation data is calculated, the correlation coefficient of the correlation data is improved, the correlation coefficient of the correlation data is optimized, and the personalized characteristic of the target data is improved.
Step A5: obtaining a target user interest monitoring analysis model by utilizing the information perception prediction end, and inputting the accurate target user personalized feature monitoring information Jb= [ J1, J2, & gt, jn ] into the target user interest monitoring analysis model for analysis to obtain target user interest recommendation data;
Specifically, the target user interest monitoring analysis model is obtained, the target user interest monitoring analysis model is a multi-level network layer, the target user interest monitoring analysis model is placed in the information perception prediction end data semantic analysis and label establishment, data semantic analysis and label establishment are carried out on the target user personalized characteristic data set, then the accurate target user personalized characteristic monitoring information Jb= [ J1, J2, ], jn ] is transmitted to the target user interest monitoring analysis model, data are transmitted to a multi-fusion model calculation component through a personalized characteristic input component, target user interest coefficient evaluation is carried out to generate target user interest recommendation coefficients, data are further transmitted to an interest recommendation analysis component, target user interest recommendation satisfaction evaluation is carried out to determine target user interest recommendation data, and the target user interest recommendation data are further output through a recommendation output component.
Further, the obtaining the target user interest monitoring analysis model by using the information perception prediction end, inputting the accurate target user personalized feature monitoring information jb= [ J1, J2 ], and Jn ] into the target user interest monitoring analysis model for analysis, and obtaining the target user interest recommendation data further includes:
The target user interest monitoring analysis model comprises a personalized feature input component, a multi-fusion model calculation component, an interest recommendation analysis component and a recommendation output component;
Inputting the personalized feature monitoring information Jb= [ J1, J2. ] of the accurate target user into the multi-fusion model calculation component through the personalized feature input component to obtain the interest recommendation coefficient of the target user;
Inputting the target user interest recommendation coefficient and the accurate target user personalized feature monitoring information Jb= [ J1, J2, ] into the interest recommendation analysis component, obtaining the target user interest recommendation data and outputting the target user interest recommendation data through the recommendation output component.
Specifically, the information perception prediction end is utilized to establish the target user interest monitoring analysis model, after semantic analysis and label establishment are carried out on the target user personalized feature data set, the target user interest monitoring analysis model is utilized to carry out data interest analysis, the target user interest monitoring analysis model is a multi-level network layer and comprises the personalized feature input component, the multi-fusion model calculation component, the interest recommendation analysis component and the recommendation output component, wherein the personalized feature input component and the recommendation output component are model necessity structures and have no special effect, the accurate target user personalized feature monitoring information Jb= [ J1, J2. ], jn ] is transmitted to the multi-fusion model calculation component through the personalized feature input component, and (3) carrying out interest coefficient evaluation on the data to obtain interest recommendation coefficients of all target users, wherein the interest recommendation coefficients of the target users are another visual expression for expressing the interest recommendation degree of the target users, further transmitting the interest recommendation coefficients of the target users and the personalized characteristic monitoring information Jb= [ J1, J2, ], jn ] of the accurate target users to the interest recommendation analysis assembly, mapping and corresponding the interest recommendation coefficients to each other, further carrying out interest recommendation analysis of the matched target users on the corresponding information, carrying out target user product analysis to obtain interest recommendation data of the target users, further outputting the interest recommendation data by using the recommendation output assembly, and carrying out interest recommendation coefficient analysis and evaluation by establishing an interest monitoring analysis model of the target users, so that objectivity of recommendation results can be ensured.
Step A6: setting an interest recommendation coefficient node value, and sending an interest adjustment signal when the interest recommendation data of the target user is less than or equal to the interest recommendation coefficient node value;
Step A7: and carrying out man-machine interaction adjustment management on the interest recommendation data of the target user by utilizing the interest adjustment signal and the self-adaptive intelligent recommendation terminal.
Specifically, target user interest recommendation analysis is performed on the accurate target user personalized feature monitoring information Jb= [ J1, J2 ], jn ] to obtain target user interest recommendation data, the target user interest recommendation coefficient corresponding to each target user is extracted and is respectively compared with the interest recommendation coefficient node value, the interest recommendation coefficient node value is a coefficient critical value for judging target user interest recommendation, whether the target user interest recommendation coefficient node value is reached or not is judged, when the interest recommendation coefficient node value is smaller than or equal to the target user interest recommendation threshold value, the fact that the target user interest recommendation satisfaction degree is insufficient is indicated, a certain target user product possibly exists, the interest adjustment signal is generated to perform adjustment warning, the target user interest analysis information is further transmitted to the self-adaptive intelligent recommendation end along with the receiving of the interest adjustment signal, the target user interest recommendation coefficient and the interest recommendation coefficient node value covariance is calculated, the target user product satisfaction degree is determined by using a covariance calculation result, a corresponding recommendation product updating scheme is further determined by using the target user product type and the product satisfaction degree, the self-adaptive intelligent recommendation end is used for performing visual display, the target user updating scheme is further guaranteed, and the target user is maintained by using the target user normally.
The man-machine interaction adjustment management of the target user interest recommendation data by using the interest adjustment signal and the self-adaptive intelligent recommendation terminal further comprises:
Acquiring recommended product type information according to the interest recommended data of the target user;
obtaining satisfaction information of a recommended product by using covariance of the interest recommendation data of the target user and the interest recommendation coefficient node value;
establishing a target user personalized feature database, and matching according to the recommended product type information, the recommended product satisfaction degree information and the target user personalized feature database to obtain a recommended product updating scheme;
And recommending and adjusting the target user by using the recommended product updating scheme.
Specifically, the target user interest monitoring analysis model is utilized to conduct interest recommendation analysis on the accurate target user personalized feature monitoring information Jb= [ J1, J2. ], jn ], the product type, the product feature and the like possibly existing in each target user are determined, the recommended product type information is obtained, the interest recommendation coefficient node value is a coefficient critical value for judging target user interest recommendation, coefficient covariance calculation is conducted on each target user in the target user interest recommendation data and the interest recommendation coefficient node value, the coefficient covariance is a product interval recommended by each target user interest, a product satisfaction judging standard is set, namely, a standard for classifying target user product satisfaction by utilizing coefficient covariance is set, for example, coefficient covariance is set to be 0.5 as a classification interval, and carrying out target user recommended product satisfaction judgment by utilizing the product satisfaction judgment standard and the target user coefficient covariance, acquiring recommended product satisfaction information, ensuring the accuracy of the product satisfaction information, establishing the target user personalized feature database, wherein the target user personalized feature database comprises maintenance schemes of each target user under different product types and different product satisfaction, carrying out suitability analysis on the recommended product type information, the recommended product satisfaction information and the target user personalized feature database, determining an update scheme corresponding to each recommended product, and further carrying out corresponding identification on the recommended product update scheme and the product target user information, so as to facilitate identification and distinguishing, and further carrying out recommendation adjustment on the target user.
With the intelligent recommendation method for sensing data in the foregoing embodiment, as shown in fig. 6, the present application provides an implementation unit of the intelligent recommendation method for sensing data, which includes:
The target user monitoring platform building unit is used for building a target user information monitoring platform, and comprises an information acquisition knowledge graph building end, an information perception prediction end and a self-adaptive intelligent recommendation end;
The target user personalized feature data set establishing unit is used for acquiring a target user personalized feature data set through the information acquisition knowledge graph establishing end, and the target user personalized feature data set comprises target user unit time preference information Qe, target user unit time behavior information Fz and target user unit time character information Tr;
The target user personalized feature data transmission unit is used for establishing a personal information knowledge graph through the information acquisition knowledge graph establishment end and then sending the personal information knowledge graph to the information perception prediction end through different factor weight models;
The semantic analysis and label establishment unit is used for carrying out semantic analysis and label establishment on the target user personalized feature data set through the information perception prediction end to obtain accurate target user personalized feature monitoring information Jb= [ J1, J2, the first place, jn ];
The user interest monitoring and analyzing unit is used for obtaining a target user interest monitoring and analyzing model by utilizing the information perception prediction end, inputting the accurate target user personalized feature monitoring information Jb= [ J1, J2, ], jn ] into the target user interest monitoring and analyzing model for analysis, and obtaining target user interest recommendation data;
the node value judging unit is used for setting an interest recommendation coefficient node value, and sending an interest adjustment signal when the interest recommendation data of the target user is less than or equal to the interest recommendation coefficient node value;
And the man-machine interaction adjustment management unit is used for carrying out man-machine interaction adjustment management on the interest recommendation data of the target user by utilizing the interest adjustment signal and the self-adaptive intelligent recommendation terminal.
Further, the method implementation further includes:
The personalized feature processing algorithm unit is used for setting a personalized feature processing algorithm, and the personalized feature processing algorithm comprises principal component analysis, whale algorithm tracking and bat algorithm local positioning;
The principal component analysis algorithm calculation unit is used for carrying out principal component analysis on the target user personalized characteristic data set to obtain target user personalized characteristic data information with correlation;
the whale algorithm tracking unit is used for tracking the whale algorithm according to the target user personalized characteristic data information with correlation to obtain a target user personalized characteristic data information boundary range;
And the bat algorithm local positioning unit is used for performing bat algorithm local positioning on the target user personalized characteristic data information boundary range by utilizing a data fluctuation range to obtain the accurate target user personalized characteristic monitoring information Jb= [ J1, J2 ].
The searching condition setting unit is used for obtaining data searching conditions, and carrying out data surrounding on the target user personalized characteristic data information with correlation by utilizing the data searching conditions to obtain effective target user personalized characteristic data information;
The coordination filtering algorithm unit is used for performing coordination filtering on the personalized characteristic data information of the effective target user to obtain the user similarity of the recommended data;
The pearson correlation coefficient calculation unit is used for performing pearson correlation coefficient calculation on the effective target user personalized feature data information when the recommended data user similarity reaches a preset data user similarity, so as to obtain correction target user personalized feature data information;
And the data information boundary range establishing unit is used for carrying out feature coding on the personalized feature data information of the correction target user to obtain the personalized feature data information boundary range of the target user.
The average score calculating unit is used for calculating the average score of the article to the recommendation correlation data information of the personalized feature data information of the effective target user to obtain the average score of the article;
The score bias calculation unit is used for obtaining a correlation data score bias when the average score of the article is larger than a preset average score;
The missing pearson correlation coefficient calculation unit is used for carrying out pearson correlation coefficient calculation on the recommended correlation data information by utilizing a data co-occurrence matrix and the correlation data scoring bias;
and the correlation data acquisition unit is used for carrying out data re-acquisition optimization on the recommended correlation data information when the average score of the article is less than or equal to the preset average score.
The distribution range searching unit is used for carrying out preset distribution range searching on the personalized characteristic data information of the effective target user by utilizing the correlation data grading bias to obtain distribution correlation data information;
the data correlation evaluation unit is used for performing correlation evaluation on the distribution correlation data information to obtain a correlation data correlation coefficient;
The correlation data ordering model building unit is used for building a correlation data ordering model according to the correlation coefficient of the correlation data and the distribution correlation data information;
And the pearson correlation coefficient optimization unit is used for calculating the pearson correlation coefficient of the recommended correlation data information by using the correlation data ordering model.
The user interest monitoring and analyzing unit is used for the target user interest monitoring and analyzing model and comprises a personalized feature input component, a multi-fusion model calculation component, an interest recommendation analysis component and a recommendation output component;
The target user interest recommendation coefficient calculation unit is used for inputting the accurate target user personalized feature monitoring information Jb= [ J1, J2 ],. Jn ] into the multi-fusion model calculation component through the personalized feature input component to obtain a target user interest recommendation coefficient;
the interest recommendation analysis unit is used for inputting the target user interest recommendation coefficient and the accurate target user personalized feature monitoring information Jb= [ J1, J2 ], the term, jn ] into the interest recommendation analysis component, obtaining target user interest recommendation data and outputting the target user interest recommendation data through the recommendation output component.
The product type acquisition unit is used for acquiring recommended product type information according to the interest recommendation data of the target user;
The product satisfaction obtaining unit is used for obtaining recommended product satisfaction information by utilizing covariance of the interest recommendation data of the target user and the interest recommendation coefficient node value;
the product updating scheme unit is used for establishing a target user personalized feature database, and matching the recommended product type information, the recommended product satisfaction degree information and the target user personalized feature database according to the recommended product type information and the recommended product satisfaction degree information to obtain a recommended product updating scheme;
And the target user adjusting unit is used for recommending and adjusting the target user by utilizing the recommended product updating scheme.
In addition, the invention also provides a computer device and a computer readable storage medium, wherein the computer device comprises a memory and a processor, the memory stores a computer program, and the processor realizes the intelligent recommendation method of the perception data when executing the computer program. The computer readable storage medium stores a computer program which when executed by a processor implements a perceived data intelligent recommendation method.
In the present specification, through the foregoing detailed description of a method, a computer device and a storage medium for intelligent recommendation of perception data, those skilled in the art can clearly understand that a method, a computer device and a storage medium for intelligent recommendation of perception data in this embodiment, for an apparatus disclosed in the embodiments, since the apparatus corresponds to a method disclosed in the embodiments, the description is relatively simple, and the relevant points refer to the description of the method section only
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (6)
1. An intelligent recommendation method for perceived data is characterized by comprising the following steps:
establishing a target user information monitoring platform, wherein the target user information monitoring platform comprises an information acquisition knowledge graph establishing end, an information perception predicting end and a self-adaptive intelligent recommending end;
acquiring a target user personalized feature data set through the information acquisition knowledge graph establishing end, wherein the target user personalized feature data set comprises target user unit time preference information Qe, target user unit time behavior information Fz and target user unit time character information Tr;
The target user unit time preference information Qe, the target user unit time behavior information Fz and the target user unit time character information Tr are transmitted to the information perception prediction end after the information acquisition knowledge graph establishment end establishes a personal information knowledge graph through different factor weight models;
The personal information knowledge graph is established through different factor weight models, and the expression is as follows:
Wherein represents a connection function between nodes of the knowledge graph, K represents a connection factor between the nodes, N represents a single node, N represents the total node number of the knowledge graph, P represents a connection path of the single node of the knowledge graph,/> represents dynamic distribution of connection paths of different single nodes, xi, ω and λ respectively represent weight coefficients of target user unit time preference information, target user unit time behavior information and target user unit time character information when the knowledge graph is constructed, and/> represents connection errors between the nodes;
semantic analysis and label establishment are carried out on the target user personalized feature data set through the information perception prediction end, so that accurate target user personalized feature monitoring information Jb= [ J1, J2., jm ] is obtained;
obtaining a target user interest monitoring analysis model by utilizing the information perception prediction end, and inputting the accurate target user personalized feature monitoring information Jb= [ J1, J2. ], jm ] into the target user interest monitoring analysis model for analysis to obtain target user interest recommendation data;
The target user interest monitoring analysis model has the expression:
Wherein represents a target user interest recommendation function, θ represents a feature factor of interest tags after semantic analysis, Y represents iteration times, Y represents a total number of target user interest categories,/> represents target user interest distribution, and/> represents a target user feature analysis error;
setting an interest recommendation coefficient node value, and sending an interest adjustment signal when the interest recommendation data of the target user is less than or equal to the interest recommendation coefficient node value;
performing man-machine interaction adjustment management on the interest recommendation data of the target user by utilizing the interest adjustment signal and the self-adaptive intelligent recommendation terminal;
the obtaining accurate target user personalized feature monitoring information jb= [ J1, J2 ], jm ], includes:
Setting a personalized feature processing algorithm, wherein the personalized feature processing algorithm comprises principal component analysis, whale algorithm tracking and bat algorithm local positioning;
Performing principal component analysis on the target user personalized feature data set to obtain target user personalized feature data information with correlation;
tracking a whale algorithm according to the target user personalized feature data information with correlation to obtain a target user personalized feature data information boundary range;
Performing bat algorithm local positioning on the target user personalized feature data information boundary range by utilizing a data fluctuation range to obtain accurate target user personalized feature monitoring information Jb= [ J1, J2 ];
The obtaining the target user personalized feature data information boundary range comprises the following steps:
acquiring data search conditions, and carrying out data surrounding on the target user personalized feature data information with correlation by utilizing the data search conditions to acquire effective target user personalized feature data information;
coordination and filtration are carried out on the personalized characteristic data information of the effective target user, and the user similarity of the recommended data is obtained;
When the recommendation data user similarity reaches the preset data user similarity, carrying out pearson correlation coefficient calculation on the effective target user personalized feature data information to obtain correction target user personalized feature data information;
Performing feature coding on the personalized feature data information of the corrected target user to obtain the boundary range of the personalized feature data information of the target user;
the pearson correlation coefficient calculation for the personalized feature data information of the effective target user comprises the following steps:
Carrying out item average score calculation on recommendation correlation data information of the personalized feature data information of the effective target user to obtain item average scores;
when the average score of the article is larger than a preset average score, obtaining a correlation data score bias;
calculating the pearson correlation coefficient of the recommended correlation data information by utilizing a data co-occurrence matrix and the correlation data scoring bias;
And when the average score of the article is less than or equal to the preset average score, carrying out data re-acquisition optimization on the recommended correlation data information.
2. The intelligent recommendation method for perceived data according to claim 1, wherein said computing pearson correlation coefficients for said recommended correlation data information using a data co-occurrence matrix and said correlation data scoring bias comprises:
Searching a preset distribution range of the personalized characteristic data information of the effective target user by utilizing the correlation data grading bias to obtain distribution correlation data information;
Carrying out correlation evaluation on the distribution correlation data information to obtain a correlation coefficient of the correlation data;
Establishing a correlation data ordering model according to the correlation coefficient of the correlation data and the distribution correlation data information;
and calculating the pearson correlation coefficient of the recommended correlation data information by using the correlation data ordering model.
3. The intelligent recommendation method for perceived data according to claim 1, wherein said obtaining target user interest recommendation data comprises:
The target user interest monitoring analysis model comprises a personalized feature input component, a multi-fusion model calculation component, an interest recommendation analysis component and a recommendation output component;
Inputting the personalized feature monitoring information Jb= [ J1, J2. ], which is the accurate target user, into the multi-fusion model calculation component through the personalized feature input component to obtain the target user interest recommendation coefficient;
Inputting the target user interest recommendation coefficient and the accurate target user personalized feature monitoring information Jb= [ J1, J2, ] into the interest recommendation analysis component, obtaining the target user interest recommendation data and outputting the target user interest recommendation data through the recommendation output component.
4. The intelligent recommendation method for perceived data according to claim 1, wherein the performing man-machine interaction adjustment management on the recommended data for interest of the target user comprises:
Acquiring recommended product type information according to the interest recommended data of the target user;
obtaining satisfaction information of a recommended product by using covariance of the interest recommendation data of the target user and the interest recommendation coefficient node value;
establishing a target user personalized feature database, and matching according to the recommended product type information, the recommended product satisfaction degree information and the target user personalized feature database to obtain a recommended product updating scheme;
And recommending and adjusting the target user by using the recommended product updating scheme.
5. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the perceived data intelligent recommendation method of any one of claims 1-4 when the computer program is executed.
6. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the intelligent recommendation method for awareness data according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310876923.2A CN116910362B (en) | 2023-07-18 | 2023-07-18 | Intelligent recommendation method for perceived data, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310876923.2A CN116910362B (en) | 2023-07-18 | 2023-07-18 | Intelligent recommendation method for perceived data, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116910362A CN116910362A (en) | 2023-10-20 |
| CN116910362B true CN116910362B (en) | 2024-04-16 |
Family
ID=88352556
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310876923.2A Active CN116910362B (en) | 2023-07-18 | 2023-07-18 | Intelligent recommendation method for perceived data, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116910362B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348895A (en) * | 2019-06-29 | 2019-10-18 | 北京淇瑀信息科技有限公司 | A kind of personalized recommendation method based on user tag, device and electronic equipment |
| CN112765486A (en) * | 2021-01-22 | 2021-05-07 | 重庆邮电大学 | Knowledge graph fused attention mechanism movie recommendation method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170169341A1 (en) * | 2015-12-14 | 2017-06-15 | Le Holdings (Beijing) Co., Ltd. | Method for intelligent recommendation |
-
2023
- 2023-07-18 CN CN202310876923.2A patent/CN116910362B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348895A (en) * | 2019-06-29 | 2019-10-18 | 北京淇瑀信息科技有限公司 | A kind of personalized recommendation method based on user tag, device and electronic equipment |
| CN112765486A (en) * | 2021-01-22 | 2021-05-07 | 重庆邮电大学 | Knowledge graph fused attention mechanism movie recommendation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116910362A (en) | 2023-10-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Insight of a fuzzy regression model | |
| Cheng | Fuzzy process control: construction of control charts with fuzzy numbers | |
| CN102053971B (en) | Recommending method and equipment for sequencing-oriented collaborative filtering | |
| CN116431931B (en) | Real-time incremental data statistical analysis method | |
| WO2002079942A2 (en) | System for visual preference determination and predictive product selection | |
| CN113157752B (en) | Scientific and technological resource recommendation method and system based on user portrait and situation | |
| Orogun et al. | Predicting consumer behaviour in digital market: a machine learning approach | |
| CN112016769B (en) | Method and device for managing relative person risk prediction and information recommendation | |
| KR20210090035A (en) | Petition System based on Petition History and Petitioner Profile Analysis | |
| CN116485020B (en) | Supply chain risk identification early warning method, system and medium based on big data | |
| CN116597939A (en) | Medicine quality control management analysis system and method based on big data | |
| CN117271905B (en) | Crowd image-based lateral demand analysis method and system | |
| CN116910362B (en) | Intelligent recommendation method for perceived data, computer equipment and storage medium | |
| CN116452212B (en) | Intelligent customer service commodity knowledge base information management method and system | |
| CN116860820A (en) | User product preference analysis method and system based on data security | |
| EP3493082A1 (en) | A method of exploring databases of time-stamped data in order to discover dependencies between the data and predict future trends | |
| Malhotra et al. | A neuro-fuzzy classifier for website quality prediction | |
| CN108335147B (en) | Data analysis method and system based on user behaviors | |
| CN113407808A (en) | Method and device for judging applicability of graph neural network model and computer equipment | |
| Abbes et al. | An efficient knowledge-based approach for random variation interpretation in NDVI time series | |
| CN118012921B (en) | Man-machine interaction data processing system for intellectual property virtual experiment | |
| Chen et al. | A time effect based collaborative filtering approach for user preference statistics and recommendation | |
| Adomavicius et al. | Stability of collaborative filtering recommendation algorithms | |
| CN117421084B (en) | Cloud platform-based intelligent software data analysis system and method | |
| CN115953724B (en) | User data analysis and management method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |