CN116662415A - Intelligent matching method and system based on data mining - Google Patents

Intelligent matching method and system based on data mining Download PDF

Info

Publication number
CN116662415A
CN116662415A CN202310934932.2A CN202310934932A CN116662415A CN 116662415 A CN116662415 A CN 116662415A CN 202310934932 A CN202310934932 A CN 202310934932A CN 116662415 A CN116662415 A CN 116662415A
Authority
CN
China
Prior art keywords
user data
data
exemplary user
evaluation
network
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.)
Granted
Application number
CN202310934932.2A
Other languages
Chinese (zh)
Other versions
CN116662415B (en
Inventor
胡志强
国忠金
李小倩
姜山
夏卫振
张甲兵
赵晓东
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong Jiabing Intelligent Technology Co ltd
Tai'an Qiyi Information Technology Co ltd
Taishan University
Original Assignee
Shandong Jiabing Intelligent Technology Co ltd
Tai'an Qiyi Information Technology Co ltd
Taishan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shandong Jiabing Intelligent Technology Co ltd, Tai'an Qiyi Information Technology Co ltd, Taishan University filed Critical Shandong Jiabing Intelligent Technology Co ltd
Priority to CN202310934932.2A priority Critical patent/CN116662415B/en
Publication of CN116662415A publication Critical patent/CN116662415A/en
Application granted granted Critical
Publication of CN116662415B publication Critical patent/CN116662415B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • 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)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Databases & Information Systems (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Probability & Statistics with Applications (AREA)
  • Fuzzy Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides an intelligent matching method and system based on data mining, and relates to the technical field of artificial intelligence. In the invention, a user data evaluation network is combined, and network optimization operation is carried out on the user correlation evaluation network so as to obtain a target correlation evaluation network corresponding to the user correlation evaluation network; extracting first user data and second user data in a user data combination to be matched; and performing user matching evaluation operation on the first user data and the second user data by using the target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination. Based on the above, the reliability of data matching can be improved to some extent.

Description

Intelligent matching method and system based on data mining
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to an intelligent matching method and system based on data mining.
Background
Data mining is used in many scenarios, for example, the characteristics of corresponding user data can be mined through a data mining operation, so that operations such as data matching analysis can be performed based on the mined characteristics. However, the conventional technique has a problem that the reliability of data matching is not high.
Disclosure of Invention
In view of the above, the present invention aims to provide an intelligent matching method and system based on data mining, so as to improve the reliability of data matching to a certain extent.
In order to achieve the above purpose, the embodiment of the present invention adopts the following technical scheme:
an intelligent matching method based on data mining, comprising:
performing network optimization operation on a user correlation evaluation network by combining a user data evaluation network to obtain a target correlation evaluation network corresponding to the user correlation evaluation network, wherein the user data evaluation network and the user correlation evaluation network belong to a neural network;
extracting first user data and second user data in a user data combination to be matched, wherein the first user data and the second user data are used for reflecting user characteristics of corresponding users to be matched, and the first user data and the second user data belong to image data;
and performing user matching evaluation operation on the first user data and the second user data by using a target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination, wherein the matching characterization data is used for reflecting whether the first user data and the second user data are matched or whether a user to be matched corresponding to the first user data and a user to be matched corresponding to the second user data are matched.
In some preferred embodiments, in the above data mining-based intelligent matching method, the step of performing, with a target matching evaluation network, a user matching evaluation operation on the first user data and the second user data to output matching characterization data between the first user data and the second user data in the user data combination includes:
respectively performing feature space mapping operation on the first user data and the second user data by using the target matching evaluation network so as to output a space mapping feature representation corresponding to the first user data and a space mapping feature representation corresponding to the second user data;
performing feature mining operation on the spatial mapping feature representation corresponding to the first user data and the spatial mapping feature representation corresponding to the second user data to form a first global feature representation comprising global related information of the first user data and a second global feature representation comprising global related information of the second user data;
performing cascading combination operation on the first global feature representation and the second global feature representation to form a corresponding cascading global feature representation;
And carrying out evaluation operation of the matching characteristic data on the cascade global characteristic representation so as to output the matching characteristic data between the first user data and the second user data in the user data combination.
In some preferred embodiments, in the above data mining-based intelligent matching method, the step of performing a network optimization operation on the user correlation evaluation network in combination with the user data evaluation network to obtain a target correlation evaluation network corresponding to the user correlation evaluation network includes:
performing a user data evaluation operation on first exemplary user data in an exemplary user data combination by using a user data evaluation network to output user evaluation data corresponding to the first exemplary user data, wherein the first exemplary user data belongs to one exemplary user data in the exemplary user data combination and is used for reflecting user characteristics of corresponding exemplary users, the first exemplary user data belongs to image data, and the user evaluation data is used for reflecting whether the first exemplary user data and second exemplary user data in the exemplary user data combination match or whether exemplary users corresponding to the first exemplary user data and the second exemplary user data match;
According to the fairness state of the exemplary user data combination cluster, performing importance analysis operation on the user evaluation data corresponding to the first exemplary user data to output an importance characterization parameter corresponding to the exemplary user data combination;
performing user matching evaluation operation on the first exemplary user data and the second exemplary user data in the exemplary user data combination by using a user matching evaluation network so as to output user matching evaluation data corresponding to the exemplary user data combination;
in case the user matching performance evaluation data corresponding to the exemplary user data combination is the occurrence probability of each actual matching performance data of the exemplary user data combination, performing a first operation on the negative correlation parameter of the occurrence probability of each actual matching performance data of the exemplary user data combination to form a corresponding target operation output parameter;
performing a second operation on the target operation output parameter and the importance characterization parameter corresponding to the exemplary user data combination to form a network optimization cost parameter corresponding to the user matching evaluation network;
and performing network optimization operation on the user matching evaluation network, so that the network optimization cost parameter is smaller than or equal to a pre-configured reference network optimization cost parameter, and obtaining a target matching evaluation network corresponding to the user matching evaluation network.
In some preferred embodiments, in the foregoing data mining-based intelligent matching method, before the step of performing, by using a user data evaluation network, a user data evaluation operation on a first exemplary user data in an exemplary user data combination to output user evaluation data corresponding to the first exemplary user data, the step of performing, by using the user data evaluation network in combination, a network optimization operation on a user relevance evaluation network to obtain a target relevance evaluation network corresponding to the user relevance evaluation network, further includes:
performing network optimization operation on candidate user data evaluation networks according to actual matching data of third exemplary user data and fourth exemplary user data in an exemplary user data combination cluster and the third exemplary user data to form a network-optimized user data evaluation network, wherein the third exemplary user data and the fourth exemplary user data belong to one exemplary user data combination in the exemplary user data combination cluster;
the step of performing a user data evaluation operation on first exemplary user data in an exemplary user data combination by using a user data evaluation network to output user evaluation data corresponding to the first exemplary user data includes:
And performing user data evaluation operation on the first exemplary user data by using the network-optimized user data evaluation network so as to output user evaluation data corresponding to the first exemplary user data.
In some preferred embodiments, in the foregoing intelligent matching method based on data mining, the step of performing network optimization operation on candidate user data evaluation networks according to actual matching data of third exemplary user data and fourth exemplary user data in the exemplary user data cluster, where the third exemplary user data is used to form a network-optimized user data evaluation network includes:
classifying the exemplary user data combined clusters to form a first number of exemplary user data combined sub-clusters corresponding to the exemplary user data combined clusters;
for any one of the first number of exemplary user data combination sub-clusters to be processed, performing the following operations:
marking the exemplary user data combination sub-clusters except the exemplary user data combination sub-cluster to be processed to be the exemplary user data combination sub-cluster to be analyzed, and performing network optimization operation on the candidate user data evaluation network by utilizing the exemplary user data combination sub-cluster to be analyzed to form a network-optimized user data evaluation network;
The step of performing a user data evaluation operation on the first exemplary user data by using the network-optimized user data evaluation network to output user evaluation data corresponding to the first exemplary user data includes:
and according to the to-be-processed exemplary user data combination sub-cluster, performing user data evaluation operation on the to-be-processed exemplary user data combination sub-cluster by utilizing the network-optimized user data evaluation network so as to output user evaluation data corresponding to the to-be-processed exemplary user data combination sub-cluster.
In some preferred embodiments, in the foregoing data mining-based intelligent matching method, the step of performing, by using the network-optimized user data evaluation network, a user data evaluation operation on the first exemplary user data to output user evaluation data corresponding to the first exemplary user data includes:
evaluating a network by utilizing the user data after network optimization, and performing feature space mapping operation on the first exemplary user data to output a space mapping feature representation corresponding to the first exemplary user data;
Performing feature mining operation on the spatial mapping feature representation corresponding to the first exemplary user data to form a global feature representation corresponding to the first exemplary user data and comprising global related information;
and carrying out user data evaluation operation on the global feature representation comprising the global related information so as to output user evaluation data corresponding to the first exemplary user data.
In some preferred embodiments, in the above-mentioned intelligent matching method based on data mining, the user evaluation data corresponding to the first exemplary user data is a likelihood of each actual matching data of the first exemplary user data;
the step of performing importance analysis operation on the user evaluation data corresponding to the first exemplary user data according to the fairness status of the exemplary user data combination cluster to output the importance characterization parameter corresponding to the exemplary user data combination includes:
analyzing fairness likelihood of each actual match data of first and second exemplary user data in the exemplary user data combination when the exemplary user data combination cluster is in a fair state;
And carrying out importance analysis operation on the user evaluation data of the first exemplary user data based on the possibility of each actual matching data of the first exemplary user data and the fairness possibility of each actual matching data so as to output an importance characterization parameter corresponding to the exemplary user data combination.
In some preferred embodiments, in the above-mentioned intelligent matching method based on data mining, the step of analyzing fairness likelihood of each actual matching data of the first exemplary user data and the second exemplary user data in the exemplary user data combination when the exemplary user data combination cluster is in a fairness state includes:
performing a round robin calculation operation on the pending fairness likelihood according to the historical likelihood of each actual match data and the likelihood of each actual match data of the first exemplary user data to form a fairness likelihood of each actual match data of the first and second exemplary user data in the exemplary user data combination, the fairness likelihood of each actual match data including a first fairness likelihood characterizing a matched first actual match data and a second fairness likelihood characterizing a non-matched second actual match data;
And the step of performing an importance analysis operation on the user evaluation data of the first exemplary user data based on the likelihood of each actual matching data of the first exemplary user data and the fairness likelihood of each actual matching data to output an importance characterization parameter corresponding to the exemplary user data combination, includes:
extracting a first likelihood of first actual match data for the first exemplary user data and extracting a second likelihood of second actual match data for the first exemplary user data, and extracting a first fair likelihood of the first actual match data and extracting a second fair likelihood of the second actual match data;
performing superposition fusion operation on a first multiplication fusion parameter between the first possibility and the second fairness possibility, and a second multiplication fusion parameter of the second possibility and the first fairness possibility to output a corresponding target superposition fusion parameter;
performing a division operation on the target superposition fusion parameter and the first multiplication fusion parameter to output an importance characterization parameter of the exemplary user data combination corresponding to the first actual matching data;
And performing a division operation on the target superposition fusion parameter and the second multiplication fusion parameter to output an importance characterization parameter of the exemplary user data combination corresponding to the second actual matching data.
In some preferred embodiments, in the above data mining-based intelligent matching method, the step of performing, by using a user matching evaluation network, a user matching evaluation operation on the first exemplary user data and the second exemplary user data in the exemplary user data combination to output user matching evaluation data corresponding to the exemplary user data combination includes:
performing feature space mapping operation on the first exemplary user data and the second exemplary user data respectively to output a space mapping feature representation corresponding to the first exemplary user data and a space mapping feature representation corresponding to the second exemplary user data;
performing feature mining operations on the spatial mapping feature representation corresponding to the first exemplary user data and the spatial mapping feature representation corresponding to the second exemplary user data to form an example first global feature representation comprising global relevant information of the first exemplary user data and to form an example second global feature representation comprising global relevant information of the second exemplary user data;
Performing cascading combination operation on the example first global feature representation and the example second global feature representation to form corresponding example cascading global feature representations;
and carrying out evaluation operation on the matching characteristic data on the example cascading global characteristic representation so as to output user matching characteristic evaluation data corresponding to the example user data combination.
The embodiment of the invention also provides an intelligent matching system based on data mining, which comprises a processor and a memory, wherein the memory is used for storing a computer program, and the processor is used for executing the computer program so as to realize the intelligent matching method based on data mining.
The intelligent matching method and the system based on data mining provided by the embodiment of the invention can be used for carrying out network optimization operation on the user correlation evaluation network by combining the user data evaluation network so as to obtain the target correlation evaluation network corresponding to the user correlation evaluation network; extracting first user data and second user data in a user data combination to be matched; and performing user matching evaluation operation on the first user data and the second user data by using the target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination. Based on the foregoing, because the user data evaluation network is combined in the process of performing the network optimization operation on the user correlation evaluation network, the accuracy of the formed target correlation evaluation network can be higher, and therefore, the reliability of data matching can be improved to a certain extent.
In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
Fig. 1 is a block diagram of an intelligent matching system based on data mining according to an embodiment of the present invention.
Fig. 2 is a schematic flow chart of steps included in the data mining-based intelligent matching method according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of each module included in the intelligent matching device based on data mining according to the embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, the embodiment of the invention provides an intelligent matching system based on data mining. The intelligent matching system based on data mining can comprise a memory and a processor.
In detail, the memory and the processor are electrically connected directly or indirectly to realize transmission or interaction of data. For example, electrical connection may be made to each other via one or more communication buses or signal lines. The memory may store at least one software functional module (computer program) that may exist in the form of software or firmware. The processor may be configured to execute the executable computer program stored in the memory, so as to implement the intelligent matching method based on data mining provided by the embodiment of the present invention.
It will be appreciated that in some possible embodiments, the Memory may be, but is not limited to, random access Memory (Random Access Memory, RAM), read Only Memory (ROM), programmable Read Only Memory (Programmable Read-Only Memory, PROM), erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), and the like.
It will be appreciated that in some possible embodiments, the processor may be a general purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), a System on Chip (SoC), etc.; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
It will be appreciated that in some possible embodiments, the intelligent matching system based on data mining may be a server with data processing capabilities.
With reference to fig. 2, the embodiment of the invention further provides an intelligent matching method based on data mining, which can be applied to the intelligent matching system based on data mining. The method steps defined by the flow related to the intelligent matching method based on the data mining can be realized by the intelligent matching system based on the data mining. The specific flow shown in fig. 2 will be described in detail.
Step S110, combining the user data evaluation network, and performing network optimization operation on the user correlation evaluation network to obtain a target correlation evaluation network corresponding to the user correlation evaluation network.
In the embodiment of the invention, the intelligent matching system based on data mining can combine the user data evaluation network to perform network optimization operation on the user correlation evaluation network so as to obtain the target correlation evaluation network corresponding to the user correlation evaluation network. The user data evaluation network and the user relevance evaluation network both belong to a neural network.
Step S120, extracting the first user data and the second user data in the user data combination to be matched.
In the embodiment of the invention, the intelligent matching system based on data mining can extract the first user data and the second user data in the user data combination to be matched. The first user data and the second user data are used for reflecting the user characteristics of the corresponding users to be matched, such as characteristics of different periods, the first user data and the second user data belong to image data, for example, the first user data and the second user data may be images of the same user, and the first user data and the second user data may not be images of the same user.
And step S130, performing user matching evaluation operation on the first user data and the second user data by using a target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination.
In the embodiment of the invention, the intelligent matching system based on data mining can utilize a target matching performance evaluation network to perform user matching performance evaluation operation on the first user data and the second user data so as to output matching performance characterization data between the first user data and the second user data in the user data combination. The matching characteristic data is used for reflecting whether the first user data and the second user data are matched or whether the user to be matched corresponding to the first user data and the user to be matched corresponding to the second user data are matched, that is, the matching characteristic data can comprise matching possibility parameters and unmatched possibility parameters. For example, for the same user, it may be used to reflect whether the user's features match at different times, and for different two users, it may be used to reflect whether the features of the two users match.
Based on the foregoing, because the user data evaluation network is combined in the process of performing the network optimization operation on the user correlation evaluation network, the accuracy of the formed target correlation evaluation network can be higher, and therefore, the reliability of data matching can be improved to a certain extent.
It may be appreciated that, in some possible embodiments, step S110 in the foregoing description, that is, the step of performing a network optimization operation on the user relevance evaluation network in conjunction with the user data evaluation network to obtain the target relevance evaluation network corresponding to the user relevance evaluation network, may further include the following detailed implementation details:
performing a user data evaluation operation on first exemplary user data in an exemplary user data combination by using a user data evaluation network to output user evaluation data corresponding to the first exemplary user data, wherein the first exemplary user data belongs to one exemplary user data in the exemplary user data combination and is used for reflecting user characteristics of corresponding exemplary users, the first exemplary user data belongs to image data, the user evaluation data is used for reflecting whether the first exemplary user data is matched with second exemplary user data in the exemplary user data combination or whether the first exemplary user data is matched with the second exemplary user data, that is, the user data evaluation network can analyze and evaluate only one exemplary user data in the exemplary user data combination to obtain user evaluation data corresponding to two exemplary user data in the exemplary user data combination, and the user evaluation data is used for analyzing and evaluating whether the first exemplary user data is matched with the second exemplary user data in the exemplary user data combination;
According to the fairness state of the exemplary user data combination cluster, performing importance analysis operation on the user evaluation data corresponding to the first exemplary user data to output an importance characterization parameter corresponding to the exemplary user data combination;
performing user matching evaluation operation on the first exemplary user data and the second exemplary user data in the exemplary user data combination by using a user matching evaluation network so as to output user matching evaluation data corresponding to the exemplary user data combination;
in case the user matching performance evaluation data corresponding to the exemplary user data combination is the occurrence probability of each actual matching performance data of the exemplary user data combination, such as a matching probability parameter and a non-matching probability parameter, performing a first operation on the negative correlation parameter of the occurrence probability of each actual matching performance data of the exemplary user data combination to form a corresponding target operation output parameter, for example, the product between the negative correlation parameter and the occurrence probability (parameter) is equal to a fixed value, such as a value of 1, the first operation may be a logarithmic operation or the like;
Performing a second operation on the target operational output parameter and the importance characterizing parameter corresponding to the exemplary user data combination to form a network optimization cost parameter corresponding to the user matching evaluation network, wherein the importance characterizing parameter corresponding to the exemplary user data combination may include the importance characterizing parameter corresponding to the first actual matching data and the importance characterizing parameter corresponding to the second actual matching data of the exemplary user data combination, and the target operational output parameter may include a first target operational output parameter corresponding to the first actual matching data and a second target operational output parameter corresponding to the second actual matching data, so that the first target operational output parameter and the second target operational output parameter may be weighted and summed based on the corresponding importance characterizing parameter to obtain the network optimization cost parameter corresponding to the user matching evaluation network;
and performing network optimization operation on the user matching evaluation network, so that the network optimization cost parameter is smaller than or equal to a pre-configured reference network optimization cost parameter, and obtaining a target matching evaluation network corresponding to the user matching evaluation network.
It may be appreciated that, in some possible embodiments, before the step of performing, by using the user data evaluation network, the user data evaluation operation on the first exemplary user data in the exemplary user data combination to output the user evaluation data corresponding to the first exemplary user data, the step of performing, by using the user data evaluation network in combination with the user data evaluation network, the network optimization operation on the user correlation evaluation network to obtain the target correlation evaluation network corresponding to the user correlation evaluation network, that is, step S110 may further include the following detailed implementation details:
according to the actual matching data of the third exemplary user data and the fourth exemplary user data in the exemplary user data combination cluster and the third exemplary user data, performing network optimization operation on the candidate user data evaluation network to form a network-optimized user data evaluation network, wherein the third exemplary user data and the fourth exemplary user data belong to one exemplary user data combination in the exemplary user data combination cluster, that is, the candidate user data evaluation network is utilized to perform analysis evaluation on the third exemplary user data to obtain corresponding evaluation data, and then, based on the difference between the evaluation data and the actual matching data, network optimization operation, such as network optimization operation along the direction of reducing the difference, can be performed on the candidate user data evaluation network to form the network-optimized user data evaluation network.
Based on the foregoing, the step of performing, by using the user data evaluation network, a user data evaluation operation on the first exemplary user data in the exemplary user data combination to output user evaluation data corresponding to the first exemplary user data may include: and performing user data evaluation operation on the first exemplary user data by using the network-optimized user data evaluation network so as to output user evaluation data corresponding to the first exemplary user data.
It may be appreciated that, in some possible embodiments, the step of performing the network optimization operation on the candidate user data evaluation network according to the actual matching data of the third exemplary user data and the fourth exemplary user data in the exemplary user data combination cluster and the third exemplary user data to form a network-optimized user data evaluation network may further include the following detailed implementation details:
classifying the exemplary user data combined clusters to form a first number of exemplary user data combined sub-clusters corresponding to the exemplary user data combined clusters, wherein the classifying operation can be random or configured according to requirements;
For any one of the first number of exemplary user data combination sub-clusters to be processed, performing the following operations:
marking the exemplary user data combination sub-clusters except the to-be-processed exemplary user data combination sub-cluster to be marked as an exemplary user data combination sub-cluster to be analyzed, and performing network optimization operation on the candidate user data evaluation network by utilizing the to-be-analyzed exemplary user data combination sub-cluster to form a network-optimized user data evaluation network.
Based on the foregoing, the step of performing, by using the network-optimized user data evaluation network, a user data evaluation operation on the first exemplary user data to output user evaluation data corresponding to the first exemplary user data includes: and according to the to-be-processed exemplary user data combination sub-cluster, performing user data evaluation operation on the to-be-processed exemplary user data combination sub-cluster by utilizing the network-optimized user data evaluation network so as to output user evaluation data corresponding to the to-be-processed exemplary user data combination sub-cluster.
That is, a portion of the data in the exemplary user data cluster is used for network optimization operations on candidate user data evaluation networks and another portion of the data is used for network optimization operations on user correlation evaluation networks.
It may be appreciated that, in some possible embodiments, the step of performing, by using the network-optimized user data evaluation network, the user data evaluation operation on the first exemplary user data to output the user evaluation data corresponding to the first exemplary user data may further include the following detailed implementation details:
evaluating a network by using the user data after network optimization, performing feature space mapping operation on the first exemplary user data to output a space mapping feature representation corresponding to the first exemplary user data, which can be referred to the related description below;
performing feature mining operation on the spatial mapping feature representation corresponding to the first exemplary user data to form a global feature representation corresponding to the first exemplary user data and including global related information, which may be referred to later related description;
And performing user data evaluation operation on the global feature representation comprising the global related information to output user evaluation data corresponding to the first exemplary user data, which can be referred to as related description hereinafter.
It may be appreciated that, in some possible embodiments, the step of performing, based on the likelihood that the user evaluation data corresponding to the first exemplary user data is the actual matching data of the first exemplary user data, performing an importance analysis operation on the user evaluation data corresponding to the first exemplary user data according to the fairness status of the exemplary user data combination cluster to output an importance characterizing parameter corresponding to the exemplary user data combination may further include the following detailed implementation details:
analyzing fairness possibility of each actual matching data of the first exemplary user data and the second exemplary user data in the exemplary user data combination when the exemplary user data combination cluster is in a fairness state, namely, reasonable and fair data distribution of the exemplary user data combination cluster;
based on the likelihood of each actual match data of the first exemplary user data and the fairness likelihood of each actual match data, performing an importance analysis operation (i.e., a comparison analysis of the likelihoods) on the user evaluation data of the first exemplary user data to output an importance characterization parameter corresponding to the exemplary user data combination.
It is to be appreciated that in some possible embodiments, the step of analyzing fairness likelihood of each actual match data of the first exemplary user data and the second exemplary user data in the exemplary user data combination when the exemplary user data combination cluster is in a fairness state may further include the following detailed implementation details:
performing a round robin calculation operation on the pending fairness likelihood according to the historical likelihood of each actual match data and the likelihood of each actual match data of the first exemplary user data to form a fairness likelihood of each actual match data of the first and second exemplary user data in the exemplary user data combination, the fairness likelihood of each actual match data including a first fairness likelihood characterizing a matched first actual match data and a second fairness likelihood characterizing a non-matched second actual match data; illustratively, a ratio of the historical likelihood of the actual match data characterizing the match to the historical likelihood of the actual match data not matching may be calculated first, and then, based on the ratio, the pending fair likelihood may be subjected to a round robin calculation operation, e.g., each actual match data has a historical likelihood corresponding to a ratio of 0.7:0.3, the likelihood of each actual match data of the first exemplary user data is 1 and 0, respectively, then based on the ratio 0.7:0.3, dividing the interval [1,0] to form a new interval [1, value of division point ], wherein the difference between 1 and the value of division point: the difference between the value of the division point and 0 is equal to the ratio 0.7:0.3, then, based on the ratio 0.7: and 0.3 dividing the new interval in the same way as the above, and circularly performing the steps until the dividing times are equal to the preset reference times, such as 2, 3, 4, 5 and the like, and then normalizing the upper limit value and the lower limit value of the finally obtained new interval [1, the value of the dividing point ] to form fairness probability of each actual matching data, wherein the sum of the fairness probability of each actual matching data is equal to 1.
It may be appreciated that, in some possible embodiments, the step of performing an importance analysis operation on the user evaluation data of the first exemplary user data based on the likelihood of each actual matching data of the first exemplary user data and the fairness likelihood of each actual matching data to output the importance characterizing parameters corresponding to the exemplary user data combination may further include the following detailed implementation details:
extracting a first likelihood of first actual match data for the first exemplary user data and extracting a second likelihood of second actual match data for the first exemplary user data, and extracting a first fair likelihood of the first actual match data and extracting a second fair likelihood of the second actual match data;
performing superposition fusion operation on a first multiplication fusion parameter between the first possibility and the second fairness possibility, the second multiplication fusion parameter between the second possibility and the first fairness possibility to output a corresponding target superposition fusion parameter, that is, multiplying the first possibility and the second fairness possibility, multiplying the second possibility and the first fairness possibility, and then adding two results obtained by multiplication, so that the corresponding target superposition fusion parameter can be obtained;
Performing a division operation on the target superposition fusion parameter and the first multiplication fusion parameter to output an importance characterization parameter of the exemplary user data combination corresponding to the first actual matching data, for example, the importance characterization parameter may be obtained by dividing the target superposition fusion parameter by the first multiplication fusion parameter;
and performing a division operation on the target superposition fusion parameter and the second multiplication fusion parameter to output an importance representation parameter of the exemplary user data combination corresponding to the second actual matching data, for example, the importance representation parameter may be obtained by dividing the target superposition fusion parameter by the second multiplication fusion parameter.
It may be appreciated that, in some possible embodiments, the step of performing, by using the user matching evaluation network, the user matching evaluation operation on the first exemplary user data and the second exemplary user data in the exemplary user data combination to output the user matching evaluation data corresponding to the exemplary user data combination may further include the following detailed implementation details:
Performing feature space mapping operation on the first exemplary user data and the second exemplary user data respectively to output a space mapping feature representation corresponding to the first exemplary user data and a space mapping feature representation corresponding to the second exemplary user data, that is, the first exemplary user data and the second exemplary user data can be mapped into feature spaces respectively to be represented in a vector form, so that the space mapping feature representation corresponding to the first exemplary user data and the space mapping feature representation corresponding to the second exemplary user data are obtained;
performing feature mining operations on the spatial mapping feature representation corresponding to the first exemplary user data and the spatial mapping feature representation corresponding to the second exemplary user data to form an example first global feature representation comprising global relevant information of the first exemplary user data and to form an example second global feature representation comprising global relevant information of the second exemplary user data;
performing a cascading combination operation on the example first global feature representation and the example second global feature representation to form a corresponding example cascading global feature representation, which may be { the example first global feature representation, the example second global feature representation }, for example;
And performing evaluation operation on the matching characteristic data on the example cascade global feature representation to output user matching characteristic evaluation data corresponding to the example user data combination, for example, performing full connection processing on the example cascade global feature representation to form corresponding example full connection vectors, and then processing the example full connection vectors through classification functions such as softmax and the like to obtain corresponding user matching characteristic evaluation data.
It may be appreciated that, in some possible embodiments, the step of performing feature mining on the spatial mapping feature representation corresponding to the first exemplary user data and the spatial mapping feature representation corresponding to the second exemplary user data to form an exemplary first global feature representation including global related information of the first exemplary user data, and forming an exemplary second global feature representation including global related information of the second exemplary user data may further include the following detailed implementation details:
performing feature mining operations on the spatial mapping feature representations corresponding to the first exemplary user data in a first direction to form first directional feature representations corresponding to the first exemplary user data, e.g., the spatial mapping feature representations corresponding to the first exemplary user data may be subjected to feature mining operations in a front-to-back order to form first directional feature representations corresponding to the first exemplary user data;
Performing a feature mining operation in a second direction on the spatial mapping feature representation corresponding to the first exemplary user data to form a second directional feature representation corresponding to the first exemplary user data, wherein the feature mining operation in the first direction is opposite to the feature mining operation in the second direction, for example, the spatial mapping feature representation corresponding to the first exemplary user data may be subjected to the feature mining operation in a back-to-front order to form a first directional feature representation corresponding to the first exemplary user data;
aggregating the first directional characteristic representation and the second directional characteristic representation to form a first global characteristic representation comprising globally relevant information of the first exemplary user data, e.g. the first directional characteristic representation and the second directional characteristic representation may be subjected to a cascading combination operation to form a first global characteristic representation, i.e. the first global characteristic representation may be { the first directional characteristic representation, the first directional characteristic representation };
performing feature mining operations on the spatial mapping feature representations corresponding to the second exemplary user data in a first direction to form third directional feature representations corresponding to the second exemplary user data, e.g., the spatial mapping feature representations corresponding to the second exemplary user data may be subjected to feature mining operations in a front-to-back order to form third directional feature representations corresponding to the second exemplary user data;
Performing feature mining operations in a second direction on the spatial mapping feature representations corresponding to the second exemplary user data to form fourth directional feature representations corresponding to the second exemplary user data, e.g., the spatial mapping feature representations corresponding to the second exemplary user data may be subjected to feature mining operations in a back-to-front order to form fourth directional feature representations corresponding to the second exemplary user data;
the third and fourth directional feature representations are aggregated to form a second global feature representation comprising globally relevant information of the second exemplary user data, e.g. the third and fourth directional feature representations may be cascade combined to form a second global feature representation, i.e. the second global feature representation may be { the third directional feature representation, the fourth directional feature representation }.
It may be appreciated that, in some possible embodiments, the step of performing the feature mining operation in the first direction on the spatial mapping feature representation corresponding to the first exemplary user data to form the first direction feature representation corresponding to the first exemplary user data may further include the following detailed implementation details:
According to corresponding image units (such as image pixels or image pixel rows, which can be configured according to actual requirements), splitting and sorting the spatial mapping feature representation corresponding to the first exemplary user data to form a corresponding local spatial mapping feature representation sequence, wherein the local spatial mapping feature representation sequence is composed of a plurality of local spatial mapping feature representations;
according to the sequence of each local space mapping feature representation in the local space mapping feature representation sequence, performing feature mining operation on each local space mapping feature representation in sequence to form a local feature representation corresponding to the local space mapping feature representation;
and carrying out ordered combination on the local feature representations corresponding to each local space mapping feature representation to form a first directional feature representation corresponding to the first exemplary user data.
It may be appreciated that, in some possible embodiments, the step of performing feature mining operation on each of the local spatial mapping feature representations sequentially according to the sequence of each of the local spatial mapping feature representations to form a local feature representation corresponding to the local spatial mapping feature representation may further include the following detailed implementation details:
For a first local spatial mapping feature representation in the sequence of local spatial mapping feature representations, taking the local spatial mapping feature representation directly as a local feature representation corresponding to the local spatial mapping feature representation;
for each other local spatial mapping feature representation other than the first local spatial mapping feature representation in the sequence of local spatial mapping feature representations, performing a focus feature analysis operation on the other local spatial mapping feature representations based on the local feature representation corresponding to the previous local spatial mapping feature representation of the other local spatial mapping feature representation to form a local focus feature representation corresponding to the other local spatial mapping feature representation, and performing a cascading combination operation on the other local spatial mapping feature representation and the local focus feature representation to form a local feature representation corresponding to the other local spatial mapping feature representation.
Wherein, it is understood that, in some possible embodiments, the step of aggregating the first directional characteristic representation and the second directional characteristic representation to form a first global characteristic representation including globally relevant information of the first exemplary user data may further include the following detailed implementation details:
Extracting a last partial feature representation (corresponding to a last image element, such as a last image pixel or image pixel row, etc.) of the first directional feature representations, the last partial feature representation including information of each of the previous partial feature representations;
extracting a first partial feature representation (corresponding to a first image element, such as a first image pixel or image pixel row, etc.) of said second directional feature representation, the first partial feature representation comprising information of each subsequent partial feature representation;
and performing a cascading combination operation on a last local feature representation in the first directional feature representation and a first local feature representation in the second directional feature representation to form a first global feature representation comprising globally relevant information of the first exemplary user data.
It may be understood that, in some possible embodiments, the step of performing a network optimization operation on the user matching evaluation network so that the network optimization cost parameter is less than or equal to a pre-configured reference network optimization cost parameter to obtain a target matching evaluation network corresponding to the user matching evaluation network further includes the following detailed implementation contents:
When the network optimization cost parameter is larger than a pre-configured reference network optimization cost parameter, analyzing a corresponding network optimization index according to the network optimization cost parameter;
and in the user matching evaluation network, performing reverse transmission operation (BP) on the network optimization index, and performing optimization adjustment on network parameters included in the user matching evaluation network in the transmission process so that the network optimization cost parameters are smaller than or equal to the pre-configured reference network optimization cost parameters, thereby forming a target matching evaluation network corresponding to the user matching evaluation network.
It may be appreciated that, in some possible embodiments, step S130 in the foregoing description, that is, the step of performing, by using the target matching evaluation network, a user matching evaluation operation on the first user data and the second user data to output matching characterization data between the first user data and the second user data in the user data combination may further include the following detailed implementation details:
using the target matching evaluation network, performing feature space mapping operation on the first user data and the second user data respectively to output a space mapping feature representation corresponding to the first user data and a space mapping feature representation corresponding to the second user data, that is, the first user data and the second user data can be mapped into feature spaces respectively to be represented in a vector form, for example, performing embedding operation to obtain the space mapping feature representation corresponding to the first user data and the space mapping feature representation corresponding to the second user data;
Performing feature mining operation on the spatial mapping feature representation corresponding to the first user data and the spatial mapping feature representation corresponding to the second user data to form a first global feature representation comprising global relevant information of the first user data and to form a second global feature representation comprising global relevant information of the second user data, as described in the previous relevance;
performing a cascading combination operation on the first global feature representation and the second global feature representation to form a corresponding cascading global feature representation, for example, the cascading global feature representation may be { the first global feature representation, the second global feature representation };
and performing evaluation operation on the cascade global feature representation to output the matching characteristic data between the first user data and the second user data in the user data combination, for example, performing full connection processing on the cascade global feature representation to form corresponding full connection vectors, and then processing the full connection vectors through classification functions such as softmax and the like to obtain the corresponding matching characteristic data.
With reference to fig. 3, the embodiment of the invention also provides an intelligent matching device based on data mining, which can be applied to the intelligent matching system based on data mining. Wherein, the intelligent matching device based on data mining can include:
the network optimization module is used for carrying out network optimization operation on the user correlation evaluation network by combining the user data evaluation network so as to obtain a target correlation evaluation network corresponding to the user correlation evaluation network, wherein the user data evaluation network and the user correlation evaluation network belong to a neural network;
the user data extraction module is used for extracting first user data and second user data in a user data combination to be matched, wherein the first user data and the second user data are used for reflecting the user characteristics of corresponding users to be matched, and the first user data and the second user data belong to image data;
the user matching evaluation module is used for performing user matching evaluation operation on the first user data and the second user data by utilizing a target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination, wherein the matching characterization data is used for reflecting whether the first user data and the second user data are matched or whether a user to be matched corresponding to the first user data and a user to be matched corresponding to the second user data are matched.
In summary, the data mining-based intelligent matching method and system provided by the invention can be used for carrying out network optimization operation on the user correlation evaluation network by combining the user data evaluation network to obtain the target correlation evaluation network corresponding to the user correlation evaluation network; extracting first user data and second user data in a user data combination to be matched; and performing user matching evaluation operation on the first user data and the second user data by using the target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination. Based on the foregoing, because the user data evaluation network is combined in the process of performing the network optimization operation on the user correlation evaluation network, the accuracy of the formed target correlation evaluation network can be higher, and therefore, the reliability of data matching can be improved to a certain extent.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. An intelligent matching method based on data mining is characterized by comprising the following steps:
performing network optimization operation on a user correlation evaluation network by combining a user data evaluation network to obtain a target correlation evaluation network corresponding to the user correlation evaluation network, wherein the user data evaluation network and the user correlation evaluation network belong to a neural network;
extracting first user data and second user data in a user data combination to be matched, wherein the first user data and the second user data are used for reflecting user characteristics of corresponding users to be matched, and the first user data and the second user data belong to image data;
and performing user matching evaluation operation on the first user data and the second user data by using a target matching evaluation network so as to output matching characterization data between the first user data and the second user data in the user data combination, wherein the matching characterization data is used for reflecting whether the first user data and the second user data are matched or whether a user to be matched corresponding to the first user data and a user to be matched corresponding to the second user data are matched.
2. The intelligent data mining-based matching method according to claim 1, wherein the step of performing a user matching evaluation operation on the first user data and the second user data using a target matching evaluation network to output matching characterization data between the first user data and the second user data in the user data combination comprises:
respectively performing feature space mapping operation on the first user data and the second user data by using the target matching evaluation network so as to output a space mapping feature representation corresponding to the first user data and a space mapping feature representation corresponding to the second user data;
performing feature mining operation on the spatial mapping feature representation corresponding to the first user data and the spatial mapping feature representation corresponding to the second user data to form a first global feature representation comprising global related information of the first user data and a second global feature representation comprising global related information of the second user data;
performing cascading combination operation on the first global feature representation and the second global feature representation to form a corresponding cascading global feature representation;
And carrying out evaluation operation of the matching characteristic data on the cascade global characteristic representation so as to output the matching characteristic data between the first user data and the second user data in the user data combination.
3. The intelligent matching method based on data mining according to claim 1, wherein the step of performing network optimization operation on the user correlation evaluation network in combination with the user data evaluation network to obtain the target correlation evaluation network corresponding to the user correlation evaluation network comprises:
performing a user data evaluation operation on first exemplary user data in an exemplary user data combination by using a user data evaluation network to output user evaluation data corresponding to the first exemplary user data, wherein the first exemplary user data belongs to one exemplary user data in the exemplary user data combination and is used for reflecting user characteristics of corresponding exemplary users, the first exemplary user data belongs to image data, and the user evaluation data is used for reflecting whether the first exemplary user data and second exemplary user data in the exemplary user data combination match or whether exemplary users corresponding to the first exemplary user data and the second exemplary user data match;
According to the fairness state of the exemplary user data combination cluster, performing importance analysis operation on the user evaluation data corresponding to the first exemplary user data to output an importance characterization parameter corresponding to the exemplary user data combination;
performing user matching evaluation operation on the first exemplary user data and the second exemplary user data in the exemplary user data combination by using a user matching evaluation network so as to output user matching evaluation data corresponding to the exemplary user data combination;
in case the user matching performance evaluation data corresponding to the exemplary user data combination is the occurrence probability of each actual matching performance data of the exemplary user data combination, performing a first operation on the negative correlation parameter of the occurrence probability of each actual matching performance data of the exemplary user data combination to form a corresponding target operation output parameter;
performing a second operation on the target operation output parameter and the importance characterization parameter corresponding to the exemplary user data combination to form a network optimization cost parameter corresponding to the user matching evaluation network;
and performing network optimization operation on the user matching evaluation network, so that the network optimization cost parameter is smaller than or equal to a pre-configured reference network optimization cost parameter, and obtaining a target matching evaluation network corresponding to the user matching evaluation network.
4. The intelligent matching method based on data mining as claimed in claim 3, wherein, before said step of using a user data evaluation network to perform a user data evaluation operation on a first exemplary user data in an exemplary user data combination to output user evaluation data corresponding to the first exemplary user data, said step of combining the user data evaluation network to perform a network optimization operation on a user correlation evaluation network to obtain a target correlation evaluation network corresponding to the user correlation evaluation network further comprises:
performing network optimization operation on candidate user data evaluation networks according to actual matching data of third exemplary user data and fourth exemplary user data in an exemplary user data combination cluster and the third exemplary user data to form a network-optimized user data evaluation network, wherein the third exemplary user data and the fourth exemplary user data belong to one exemplary user data combination in the exemplary user data combination cluster;
the step of performing a user data evaluation operation on first exemplary user data in an exemplary user data combination by using a user data evaluation network to output user evaluation data corresponding to the first exemplary user data includes:
And performing user data evaluation operation on the first exemplary user data by using the network-optimized user data evaluation network so as to output user evaluation data corresponding to the first exemplary user data.
5. The intelligent matching method based on data mining according to claim 4, wherein the step of performing network optimization operation on candidate user data evaluation networks according to actual matching data of third exemplary user data and fourth exemplary user data in the exemplary user data cluster, the third exemplary user data, to form a network-optimized user data evaluation network comprises:
classifying the exemplary user data combined clusters to form a first number of exemplary user data combined sub-clusters corresponding to the exemplary user data combined clusters;
for any one of the first number of exemplary user data combination sub-clusters to be processed, performing the following operations:
marking the exemplary user data combination sub-clusters except the exemplary user data combination sub-cluster to be processed to be the exemplary user data combination sub-cluster to be analyzed, and performing network optimization operation on the candidate user data evaluation network by utilizing the exemplary user data combination sub-cluster to be analyzed to form a network-optimized user data evaluation network;
The step of performing a user data evaluation operation on the first exemplary user data by using the network-optimized user data evaluation network to output user evaluation data corresponding to the first exemplary user data includes:
and according to the to-be-processed exemplary user data combination sub-cluster, performing user data evaluation operation on the to-be-processed exemplary user data combination sub-cluster by utilizing the network-optimized user data evaluation network so as to output user evaluation data corresponding to the to-be-processed exemplary user data combination sub-cluster.
6. The intelligent matching method based on data mining according to claim 4, wherein the step of performing a user data evaluation operation on the first exemplary user data by using the network-optimized user data evaluation network to output user evaluation data corresponding to the first exemplary user data includes:
evaluating a network by utilizing the user data after network optimization, and performing feature space mapping operation on the first exemplary user data to output a space mapping feature representation corresponding to the first exemplary user data;
Performing feature mining operation on the spatial mapping feature representation corresponding to the first exemplary user data to form a global feature representation corresponding to the first exemplary user data and comprising global related information;
and carrying out user data evaluation operation on the global feature representation comprising the global related information so as to output user evaluation data corresponding to the first exemplary user data.
7. The intelligent matching method based on data mining according to claim 3, wherein the user evaluation data corresponding to the first exemplary user data is a likelihood of each actual matching data of the first exemplary user data;
the step of performing importance analysis operation on the user evaluation data corresponding to the first exemplary user data according to the fairness status of the exemplary user data combination cluster to output the importance characterization parameter corresponding to the exemplary user data combination includes:
analyzing fairness likelihood of each actual match data of first and second exemplary user data in the exemplary user data combination when the exemplary user data combination cluster is in a fair state;
And carrying out importance analysis operation on the user evaluation data of the first exemplary user data based on the possibility of each actual matching data of the first exemplary user data and the fairness possibility of each actual matching data so as to output an importance characterization parameter corresponding to the exemplary user data combination.
8. The intelligent data mining-based matching method according to claim 7, wherein the step of analyzing fairness likelihood of each actual match data of the first and second exemplary user data in the exemplary user data combination when the exemplary user data combination cluster is in a fairness state comprises:
performing a round robin calculation operation on the pending fairness likelihood according to the historical likelihood of each actual match data and the likelihood of each actual match data of the first exemplary user data to form a fairness likelihood of each actual match data of the first and second exemplary user data in the exemplary user data combination, the fairness likelihood of each actual match data including a first fairness likelihood characterizing a matched first actual match data and a second fairness likelihood characterizing a non-matched second actual match data;
And the step of performing an importance analysis operation on the user evaluation data of the first exemplary user data based on the likelihood of each actual matching data of the first exemplary user data and the fairness likelihood of each actual matching data to output an importance characterization parameter corresponding to the exemplary user data combination, includes:
extracting a first likelihood of first actual match data for the first exemplary user data and extracting a second likelihood of second actual match data for the first exemplary user data, and extracting a first fair likelihood of the first actual match data and extracting a second fair likelihood of the second actual match data;
performing superposition fusion operation on a first multiplication fusion parameter between the first possibility and the second fairness possibility, and a second multiplication fusion parameter of the second possibility and the first fairness possibility to output a corresponding target superposition fusion parameter;
performing a division operation on the target superposition fusion parameter and the first multiplication fusion parameter to output an importance characterization parameter of the exemplary user data combination corresponding to the first actual matching data;
And performing a division operation on the target superposition fusion parameter and the second multiplication fusion parameter to output an importance characterization parameter of the exemplary user data combination corresponding to the second actual matching data.
9. The intelligent matching method based on data mining as claimed in claim 3, wherein said step of performing a user matching evaluation operation on the first exemplary user data and the second exemplary user data in the exemplary user data combination using the user matching evaluation network to output user matching evaluation data corresponding to the exemplary user data combination comprises:
performing feature space mapping operation on the first exemplary user data and the second exemplary user data respectively to output a space mapping feature representation corresponding to the first exemplary user data and a space mapping feature representation corresponding to the second exemplary user data;
performing feature mining operations on the spatial mapping feature representation corresponding to the first exemplary user data and the spatial mapping feature representation corresponding to the second exemplary user data to form an example first global feature representation comprising global relevant information of the first exemplary user data and to form an example second global feature representation comprising global relevant information of the second exemplary user data;
Performing cascading combination operation on the example first global feature representation and the example second global feature representation to form corresponding example cascading global feature representations;
and carrying out evaluation operation on the matching characteristic data on the example cascading global characteristic representation so as to output user matching characteristic evaluation data corresponding to the example user data combination.
10. A data mining based intelligent matching system comprising a processor and a memory, the memory for storing a computer program, the processor for executing the computer program to implement the data mining based intelligent matching method of any of claims 1-9.
CN202310934932.2A 2023-07-28 2023-07-28 Intelligent matching method and system based on data mining Active CN116662415B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310934932.2A CN116662415B (en) 2023-07-28 2023-07-28 Intelligent matching method and system based on data mining

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310934932.2A CN116662415B (en) 2023-07-28 2023-07-28 Intelligent matching method and system based on data mining

Publications (2)

Publication Number Publication Date
CN116662415A true CN116662415A (en) 2023-08-29
CN116662415B CN116662415B (en) 2023-10-20

Family

ID=87724527

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310934932.2A Active CN116662415B (en) 2023-07-28 2023-07-28 Intelligent matching method and system based on data mining

Country Status (1)

Country Link
CN (1) CN116662415B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020038254A1 (en) * 2018-08-23 2020-02-27 杭州海康威视数字技术股份有限公司 Image processing method and apparatus for target recognition
US20200151649A1 (en) * 2018-11-14 2020-05-14 Servicenow, Inc. Workflows with rule-based assignments
CN112749758A (en) * 2021-01-21 2021-05-04 北京百度网讯科技有限公司 Image processing method, neural network training method, device, equipment and medium
CN116091796A (en) * 2023-04-11 2023-05-09 天之翼(苏州)科技有限公司 Unmanned aerial vehicle acquisition data processing method and system based on artificial intelligence
CN116109630A (en) * 2023-04-10 2023-05-12 创域智能(常熟)网联科技有限公司 Image analysis method and system based on sensor acquisition and artificial intelligence
CN116126945A (en) * 2023-03-30 2023-05-16 创域智能(常熟)网联科技有限公司 Sensor running state analysis method and system based on data analysis

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020038254A1 (en) * 2018-08-23 2020-02-27 杭州海康威视数字技术股份有限公司 Image processing method and apparatus for target recognition
US20200151649A1 (en) * 2018-11-14 2020-05-14 Servicenow, Inc. Workflows with rule-based assignments
CN112749758A (en) * 2021-01-21 2021-05-04 北京百度网讯科技有限公司 Image processing method, neural network training method, device, equipment and medium
CN116126945A (en) * 2023-03-30 2023-05-16 创域智能(常熟)网联科技有限公司 Sensor running state analysis method and system based on data analysis
CN116109630A (en) * 2023-04-10 2023-05-12 创域智能(常熟)网联科技有限公司 Image analysis method and system based on sensor acquisition and artificial intelligence
CN116091796A (en) * 2023-04-11 2023-05-09 天之翼(苏州)科技有限公司 Unmanned aerial vehicle acquisition data processing method and system based on artificial intelligence

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
CHEN YUTIAN: "Efficient Visual Place Recognition by Adaptive CNN Landmark Matching", KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS, pages 4084 - 4104 *
杨锐雄;韦劭高;李冉;: "基于互联网+的"众筹式"网络优化模式研究", 邮电设计技术, no. 02, pages 12 - 18 *

Also Published As

Publication number Publication date
CN116662415B (en) 2023-10-20

Similar Documents

Publication Publication Date Title
CN109918498B (en) Problem warehousing method and device
CN111723226B (en) Information management method based on big data and Internet and artificial intelligence cloud server
CN116091796B (en) Unmanned aerial vehicle acquisition data processing method and system based on artificial intelligence
CN116126947B (en) Big data analysis method and system applied to enterprise management system
CN116109630B (en) Image analysis method and system based on sensor acquisition and artificial intelligence
CN115687674A (en) Big data demand analysis method and system serving smart cloud service platform
CN115294397A (en) Classification task post-processing method, device, equipment and storage medium
CN116070149A (en) Data analysis method and system based on artificial intelligence and cloud platform
CN116662415B (en) Intelligent matching method and system based on data mining
CN116821777A (en) Novel basic mapping data integration method and system
CN115757900A (en) User demand analysis method and system applying artificial intelligence model
CN110929118B (en) Network data processing method, device, apparatus and medium
CN111401392B (en) Cluster integration method and device, electronic equipment and storage medium
CN115908119B (en) Face image beautifying processing method and system based on artificial intelligence
CN116628181B (en) User control preference sensing method and system based on Internet of things
CN116955831A (en) Information recommendation method and system based on data mining and AI model analysis
CN116680323B (en) User demand mining method and system based on big data security platform
CN116738396B (en) Artificial intelligence-based landmark quasi document input method and system
CN117041121B (en) Internet of Things anomaly monitoring method and system based on data mining
CN115082709B (en) Remote sensing big data processing method, system and cloud platform
CN114625747B (en) Wind control updating method and system based on information security
CN111625672B (en) Image processing method, image processing device, computer equipment and storage medium
CN115620031B (en) Natural resource right-determining registration information processing method, system and equipment
CN116955451A (en) Intelligent mapping data management method and system
CN117496419A (en) Data processing method and system based on intelligent building

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