CN116109121B - User demand mining method and system based on big data analysis - Google Patents

User demand mining method and system based on big data analysis Download PDF

Info

Publication number
CN116109121B
CN116109121B CN202310404517.6A CN202310404517A CN116109121B CN 116109121 B CN116109121 B CN 116109121B CN 202310404517 A CN202310404517 A CN 202310404517A CN 116109121 B CN116109121 B CN 116109121B
Authority
CN
China
Prior art keywords
user
mining
distribution
network
data
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
Application number
CN202310404517.6A
Other languages
Chinese (zh)
Other versions
CN116109121A (en
Inventor
马衍阳
杨春晨
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.)
Xichang College
Original Assignee
Xichang College
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 Xichang College filed Critical Xichang College
Priority to CN202310404517.6A priority Critical patent/CN116109121B/en
Publication of CN116109121A publication Critical patent/CN116109121A/en
Application granted granted Critical
Publication of CN116109121B publication Critical patent/CN116109121B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06315—Needs-based resource requirements planning or analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking
    • 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

Abstract

The invention provides a user demand mining method and system based on big data analysis, and relates to the technical field of artificial intelligence. In the invention, a user distribution relation network to be analyzed is constructed; analyzing a user group distribution characterization vector corresponding to a user group in a user distribution relation network to be analyzed, and analyzing user sub-group type data corresponding to the user group distribution characterization vector; generating group user mining parameter distribution based on the group user data characterization vector of the user sub-group species data and the user distribution relation network to be analyzed; based on the user sub-group types and the distributed coordinate data, corresponding relevant users are determined; and carrying out user demand mining operation based on the user attribute data of each differential user and the user attribute data of the related user of the differential user respectively to obtain a user demand mining result corresponding to each differential user. Based on the above, the reliability of user demand mining can be improved.

Description

User demand mining method and system based on big data analysis
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a user demand mining method and system based on big data analysis.
Background
Management (also known as management science, english: management Science) is a science of studying human management activities and their applications. It is biased to solve the management problems with some tools and methods, such as quantitative qualitative analysis with operations theory, statistics, etc. Management is defined as the process by which the manager and others and through others efficiently and effectively complete an activity. Since all organizations can be considered a certain system, management can also be considered a human behavioral phenomenon, including designing, promoting better production of the system. This view creates a development opportunity for the "management" itself.
Artificial intelligence (Artificial Intelligence, AI for short) is a theory, method, technique and application system that simulates, extends and extends human intelligence, senses environment, obtains knowledge and uses knowledge to obtain optimal results using digital computers or digital computer controlled computations.
In applications of management, user demand mining based on data analysis is included, wherein in order to ensure the reliability of data analysis, data analysis can be performed based on artificial intelligence technology, however, in the prior art, there is a problem that the reliability of user demand mining is not good.
Disclosure of Invention
In view of the above, the present invention aims to provide a user demand mining method and system based on big data analysis, so as to improve the reliability of user demand mining.
In order to achieve the above purpose, the embodiment of the present invention adopts the following technical scheme:
a user demand mining method based on big data analysis comprises the following steps:
constructing a user distribution relation network to be analyzed based on the users to be analyzed;
analyzing a user group distribution characterization vector corresponding to a user group in the user distribution relation network to be analyzed by utilizing a target user mining network, and analyzing user sub-group type data corresponding to the user group distribution characterization vector, wherein the user group distribution characterization vector is used for characterizing the distribution relation of the user group, and the user sub-group type data is used for reflecting the user sub-group types corresponding to the differential users of the user group;
generating group user mining parameter distribution based on the group user data characterization vector of the user sub-group type data and the user distribution relation network to be analyzed, wherein the group user mining parameter distribution is used for reflecting distribution coordinate data of the differential users in the user distribution relation network to be analyzed;
For each of the differential users, determining relevant users corresponding to the differential users in the user distribution relation network to be analyzed based on the user sub-group types corresponding to the differential users and the distribution coordinate data of the differential users in the user distribution relation network to be analyzed;
and carrying out user demand mining operation based on the user attribute data of each differential user and the user attribute data of the related user of the differential user respectively to obtain a user demand mining result corresponding to each differential user, wherein the user demand mining result is used for reflecting the information of interest of the differential user in the form of text data.
In some preferred embodiments, in the method for mining user requirements based on big data analysis, the step of constructing a user distribution relation network to be analyzed based on the user to be analyzed includes:
determining a plurality of users to be analyzed;
respectively extracting user attribute data of each user to be analyzed, wherein the user attribute data at least comprises static attribute data and dynamic attribute data of the user to be analyzed, the static attribute data at least comprises identity data, the dynamic attribute data at least comprises behavior data, and the user attribute data at least comprises text data;
And based on the user attribute data, determining user correlation among the users to be analyzed, and based on the user correlation, performing construction operation on the relationship networks of the users to be analyzed to form a corresponding user distribution relationship network to be analyzed, wherein the distribution relationship among the users to be analyzed is related to the user correlation in the user distribution relationship network to be analyzed.
In some preferred embodiments, in the above method for mining user requirements based on big data analysis, the step of mining a network by using a target user, analyzing a user population distribution characterization vector corresponding to a user population in the user distribution relation network to be analyzed, and analyzing user sub-population category data corresponding to the user population distribution characterization vector includes:
loading the user distribution relation network to be analyzed to the target user mining network, mining an overall group characterization vector corresponding to the user group in the user distribution relation network to be analyzed by using the target user mining network, and analyzing overall group type data corresponding to the overall group characterization vector by using a type analysis unit in the target user mining network, wherein the overall group type data is used for reflecting whether a user belongs to a user group;
Extracting group user data characterization vectors which are extracted by a designated filtering unit in the target user mining network and aim at the user distribution relation network to be analyzed, and carrying out multiplication calculation operation on the whole group type data and the group user data characterization vectors so as to output whole group mining parameter distribution corresponding to the user distribution relation network to be analyzed;
dividing the user distribution relation network to be analyzed based on the overall group mining parameter distribution to form a first number of sub-group distribution relation networks, and mining sub-group representative vectors corresponding to the first number of sub-group distribution relation networks based on the target user mining network;
performing aggregation operation on the overall group characterization vector and subgroup characterization vectors corresponding to the first number of subgroup distribution relation networks to form corresponding user group distribution characterization vectors;
and analyzing the user sub-group type data corresponding to the user group distribution characterization vector.
In some preferred embodiments, in the foregoing big data analysis based user demand mining method, the big data analysis based user demand mining method further includes:
Extracting a typical user distribution relation network;
analyzing original user mining data corresponding to the typical user distribution relation network by using a candidate user mining network, and determining original user mining parameter distribution based on the original user mining data and typical user data characterization vectors of the typical user distribution relation network, wherein the original user mining data is obtained based on typical group distribution characterization vectors corresponding to typical user groups in the typical user distribution relation network, the typical group distribution characterization vectors are used for characterizing the distribution relation of user groups, and the original user mining parameter distribution is used for reflecting distribution coordinate data of differential users of the typical user groups in the typical user distribution relation network;
performing relation network adjustment operation on the typical user distribution relation network to form an adjustment user distribution relation network, analyzing adjustment user mining data corresponding to the adjustment user distribution relation network by utilizing the candidate user mining network, and determining adjustment user mining parameter distribution based on the adjustment user mining data and adjustment user data characterization vectors of the adjustment user distribution relation network, wherein the adjustment user mining data is obtained based on adjustment group distribution characterization vectors corresponding to typical user groups in the adjustment user distribution relation network, the adjustment group distribution characterization vectors are used for characterizing the distribution relation of user groups, and the adjustment user mining parameter distribution is used for reflecting distribution coordinate data of differential users of the typical user groups in the adjustment user distribution relation network;
Analyzing associated mining error indexes corresponding to the candidate user mining network based on the original user mining parameter distribution and the adjusted user mining parameter distribution;
analyzing target mining error indexes corresponding to the candidate user mining network based on the original user mining data, the adjusted user mining data and the differential user identification data corresponding to the typical user distribution relation network;
and carrying out network optimization operation on the candidate user mining network according to the associated mining error index and the target mining error index to form a target user mining network, wherein the target user mining network is used for analyzing user sub-group category information and distribution coordinate analysis data corresponding to the differential users in the user distribution relation network to be analyzed.
In some preferred embodiments, in the foregoing big data analysis-based user demand mining method, the step of using a candidate user mining network to analyze original user mining data corresponding to the typical user distribution relation network, and determining an original user mining parameter distribution based on the original user mining data and a typical user data characterization vector of the typical user distribution relation network includes:
Loading the typical user distribution relation network to load into the candidate user mining network, and mining typical group distribution characterization vectors corresponding to the typical user groups in the typical user distribution relation network by utilizing the candidate user mining network;
analyzing the representative group distribution characterization vector based on a category analysis unit in the candidate user mining network so as to output original user mining data corresponding to the representative user distribution relation network;
extracting typical user data characterization vectors which are extracted by a designated filtering unit in the candidate user mining network and aim at the typical user distribution relation network, and carrying out multiplication calculation operation on the original user mining data and the typical user data characterization vectors so as to output undetermined user mining parameter distribution corresponding to the typical user distribution relation network;
and carrying out interpolation operation on the undetermined user mining parameter distribution to form an original user mining parameter score with the same size as the typical user distribution relation network.
In some preferred embodiments, in the foregoing method for mining user requirements based on big data analysis, the step of loading the typical user distribution relation network to load into the candidate user mining network, and mining a typical population distribution characterization vector corresponding to the typical user population in the typical user distribution relation network by using the candidate user mining network includes:
Loading the typical user distribution relation network to load into the candidate user mining network;
digging an overall group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network by utilizing the candidate user mining network;
analyzing the overall user mining data corresponding to the overall group distribution characterization vector by using a category analysis unit in the candidate user mining network;
performing multiplication calculation operation on the whole user mining data and the typical user data characterization vector to output whole user mining parameter distribution corresponding to the typical user distribution relation network;
dividing the representative user distribution relationship network based on the overall user mining parameter distribution to form a first number of representative sub-population relationship networks;
loading the first number of typical sub-population relation networks respectively to load the first number of typical sub-population relation networks into the candidate user mining network, and mining sub-population distribution characterization vectors corresponding to the first number of typical sub-population relation networks respectively by utilizing the candidate user mining network;
and performing aggregation operation on the overall population distribution characterization vector and the sub-population distribution characterization vectors corresponding to the first number of typical sub-population relationship networks to form typical population distribution characterization vectors corresponding to the typical user populations in the corresponding typical user distribution relationship networks.
In some preferred embodiments, in the user demand mining method based on big data analysis described above, the candidate user mining network includes a second number of gradient optimization units, each gradient optimization unit including one or more filtering subunits;
the step of mining the overall group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network by using the candidate user mining network includes:
determining data to be processed of any one gradient optimizing unit in the second number of gradient optimizing units; in the case that any one gradient optimizing unit is the first gradient optimizing unit, the data to be processed of the any one gradient optimizing unit is the typical user distribution relation network;
filtering the data to be processed of any one gradient optimizing unit based on one or more filtering subunits in the any one gradient optimizing unit to form an intermediate filtering characterization vector;
performing aggregation operation on the intermediate filtering characterization vector and the data to be processed of any one gradient optimization unit to form optimized output data of the any one gradient optimization unit, and performing marking operation on the optimized output data of the any one gradient optimization unit to mark the optimized output data as the data to be processed of the latter gradient optimization unit;
And marking the optimized output data of the last gradient optimizing unit to be marked as an integral group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network.
In some preferred embodiments, in the user demand mining method based on big data analysis, the number of the overall population distribution characterization vectors is a third number;
the step of analyzing the overall user mining data corresponding to the overall group distribution characterization vector by using a category analysis unit in the candidate user mining network includes:
respectively calculating the average value vector parameters corresponding to the third number of the overall population distribution characterization vectors, and carrying out vector construction operation on the average value vector parameters corresponding to the third number of the overall population distribution characterization vectors to form corresponding overall average value characterization vectors;
performing excitation mapping operation on the whole mean value representation vector based on an excitation mapping unit included in the candidate user mining network so as to form a corresponding excitation mapping feature vector;
and loading the excitation mapping feature vector to load into a category analysis unit included in the candidate user mining network, and analyzing the whole user mining data corresponding to the excitation mapping feature vector by using the category analysis unit in the candidate user mining network.
In some preferred embodiments, in the above method for mining user requirements based on big data analysis, the step of performing a network optimization operation on the candidate user mining network according to the associated mining error index and the target mining error index to form a target user mining network includes:
calculating the total mining error index corresponding to the candidate user mining network based on the associated mining error index and the target mining error index;
and optimally adjusting the parameters of the candidate user mining network along the direction of reducing the total mining error index so as to form a target user mining network comprising the optimally adjusted parameters.
The embodiment of the invention also provides a user demand mining system based on big data analysis, 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 user demand mining method based on big data analysis.
The user demand mining method and system based on big data analysis provided by the embodiment of the invention can construct a user distribution relation network to be analyzed; analyzing a user group distribution characterization vector corresponding to a user group in a user distribution relation network to be analyzed, and analyzing user sub-group type data corresponding to the user group distribution characterization vector; generating group user mining parameter distribution based on the group user data characterization vector of the user sub-group species data and the user distribution relation network to be analyzed; based on the user sub-group types and the distributed coordinate data, corresponding relevant users are determined; and carrying out user demand mining operation based on the user attribute data of each differential user and the user attribute data of the related user of the differential user respectively to obtain a user demand mining result corresponding to each differential user. Based on the foregoing, in the process of performing the user demand mining operation, not only the user attribute data of the corresponding user is analyzed, but also the corresponding user attribute data of the related user is combined to perform the analysis, so that the analysis basis is more sufficient and more constraint is provided, and therefore, the reliability of the user demand mining can be improved, and the problem of low reliability in the prior art is solved.
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 a user demand mining system based on big data analysis according to an embodiment of the present invention.
Fig. 2 is a flowchart illustrating steps included in a user demand mining method based on big data analysis according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of each module included in the user demand mining apparatus based on big data analysis 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 a user demand mining system based on big data analysis. The user demand mining system may include, among other things, 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 user requirement mining method based on big data analysis provided by the embodiment of the present invention.
It is to be appreciated that in some 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 (ErasableProgrammable 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 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 embodiments, the big data analysis based user demand mining system may be a server with data processing capabilities.
With reference to fig. 2, the embodiment of the invention further provides a user demand mining method based on big data analysis, which can be applied to the user demand mining system based on big data analysis. The method steps defined by the flow related to the user demand mining method based on big data analysis can be realized by the user demand mining system based on big data analysis.
The specific flow shown in fig. 2 will be described in detail.
And step S110, constructing a user distribution relation network to be analyzed based on the user to be analyzed.
In the embodiment of the invention, the user demand mining system based on big data analysis can construct a user distribution relation network to be analyzed based on the users to be analyzed.
Step S120, a target user is utilized to mine a network, a user population distribution characterization vector corresponding to a user population in the user distribution relation network to be analyzed is analyzed, and user sub-population type data corresponding to the user population distribution characterization vector is analyzed.
In the embodiment of the invention, the user demand mining system based on big data analysis can utilize a target user mining network to analyze a user population distribution characterization vector corresponding to a user population in the user distribution relation network to be analyzed, and analyze user sub-population category data corresponding to the user population distribution characterization vector. The user population distribution characterization vector is used for characterizing the distribution relation of the user population, the user sub-population type data is used for reflecting the user sub-population types corresponding to the differential users of the user population, the differential users can be the users at the population distribution edge of the user population and can distinguish the user population from other users, in addition, the user sub-population can be different sub-populations in the user population, for example, in one large population, according to the difference between the users, the different sub-populations can be classified, for example, the user population can be the user combination with specific behavior tendency, the sub-population can be defined based on small user combinations formed by dividing different tendency degrees of specific behaviors, and the sub-population can be defined based on other modes.
And step S130, generating group user mining parameter distribution based on the user sub-group type data and the group user data characterization vector of the user distribution relation network to be analyzed.
In the embodiment of the invention, the user demand mining system based on big data analysis can generate group user mining parameter distribution based on the group user data characterization vector of the user sub-group category data and the user distribution relation network to be analyzed. And the group user mining parameter distribution is used for reflecting distribution coordinate data of the differential users in the user distribution relation network to be analyzed.
Step S140, for each of the differential users, determining, in the user distribution relationship network to be analyzed, a relevant user corresponding to the differential user based on the user sub-group category corresponding to the differential user and the distribution coordinate data of the differential user in the user distribution relationship network to be analyzed.
In the embodiment of the present invention, the user demand mining system based on big data analysis may determine, for each of the differential users, a relevant user corresponding to the differential user in the user distribution relationship network to be analyzed based on a user sub-group category corresponding to the differential user and distribution coordinate data of the differential user in the user distribution relationship network to be analyzed.
Step S150, user demand mining operation is performed based on the user attribute data of each differential user and the user attribute data of the related user of the differential user, so as to obtain a user demand mining result corresponding to each differential user.
In the embodiment of the present invention, the user demand mining system based on big data analysis may perform user demand mining operations based on user attribute data of each of the differential users and user attribute data of related users of the differential users, so as to obtain user demand mining results corresponding to each of the differential users. The user demand mining results are used to reflect (or describe) the information of interest of the diverse users in the form of text data. For example, feature mining may be performed on the user attribute data of the differential user (e.g., implemented through a coding network) to obtain a corresponding first feature mining vector, feature mining may be performed on the user attribute data of the related user to obtain a corresponding second feature mining vector, attention operation may be performed on the first feature mining vector based on the second feature mining vector to form a corresponding attention vector, then the attention vector and the first feature mining vector may be spliced (i.e., a cascade combination operation) to form a corresponding spliced vector, and then user demand prediction may be performed on the spliced vector to output a user demand mining result corresponding to the differential user.
Based on the foregoing, in the process of performing the user demand mining operation, not only the user attribute data of the corresponding user is analyzed, but also the corresponding user attribute data of the related user is combined to perform the analysis, so that the analysis basis is more sufficient and more constraint is provided, and therefore, the reliability of the user demand mining can be improved, and the problem of low reliability in the prior art is solved.
It will be appreciated that, in some embodiments, the step S110 described above, that is, the step of constructing the user distribution relationship network to be analyzed based on the user to be analyzed, further includes the following steps:
determining a plurality of users to be analyzed;
respectively extracting user attribute data of each user to be analyzed, wherein the user attribute data at least comprises static attribute data and dynamic attribute data of the user to be analyzed, the static attribute data at least comprises identity data, the dynamic attribute data at least comprises behavior data, and the user attribute data at least comprises text data;
based on the user attribute data, determining a user correlation relationship between users to be analyzed, for example, determining a relationship closeness representing the user correlation relationship based on the similarity between the user attribute data, wherein the relationship closeness is positively correlated with the similarity, and based on the user correlation relationship, performing a construction operation on a relationship network of the plurality of users to be analyzed to form a corresponding user distribution relationship network to be analyzed, wherein in the user distribution relationship network to be analyzed, the distribution relationship between users to be analyzed is correlated with the user correlation relationship, for example, the closer the user correlation relationship is, the smaller the distribution coordinate distance between the corresponding users can be, in addition, the user distribution relationship network to be analyzed can refer to two-dimensional space distribution, or three-dimensional, four-dimensional space distribution, wherein in the user distribution relationship network to be analyzed, the user attribute data can be directly used as user identification, or the user attribute data can be processed, namely, different user attribute data correspond to different numerical values, and then the numerical values can be used as the numerical values.
It may be appreciated that, in some embodiments, the step S120 of analyzing the user population distribution characterization vector corresponding to the user population in the user distribution relationship network to be analyzed by using the target user to mine the network, and the step of analyzing the user sub-population category data corresponding to the user population distribution characterization vector may further include the following:
loading the user distribution relation network to be analyzed to the target user mining network, mining an overall group characterization vector corresponding to the user group in the user distribution relation network to be analyzed by using the target user mining network, and analyzing overall group type data corresponding to the overall group characterization vector by using a type analysis unit in the target user mining network, wherein the overall group type data is used for reflecting whether a user belongs to a user group;
extracting a group user data characterization vector which is extracted by a designated filter unit in the target user mining network and aims at the user distribution relation network to be analyzed, and carrying out multiplication calculation operation on the whole group type data and the group user data characterization vector so as to output whole group mining parameter distribution corresponding to the user distribution relation network to be analyzed, wherein the designated filter unit can be the last filter unit in the target user mining network;
Dividing the user distribution relation network to be analyzed based on the overall group mining parameter distribution to form a first number of sub-group distribution relation networks, namely local relation networks corresponding to each sub-group in the user group, and mining sub-group identification vectors corresponding to each of the first number of sub-group distribution relation networks based on the target user mining network;
performing aggregation operation on the overall group characterization vector and subgroup characterization vectors corresponding to the first number of subgroup distribution relation networks to form corresponding user group distribution characterization vectors, for example, performing cascade combination operation on the overall group characterization vector and subgroup characterization vectors corresponding to the first number of subgroup distribution relation networks;
and analyzing the user sub-population type data corresponding to the user population distribution characterization vector, for example, predicting and outputting based on the user population distribution characterization vector to obtain user sub-population type data, wherein the user sub-population type data is used for reflecting the user sub-population types corresponding to the differential users.
It may be appreciated that, in some embodiments, step S140 described above, that is, the step of determining, for each of the differential users, the relevant user corresponding to the differential user in the user distribution relationship network to be analyzed based on the user sub-group category corresponding to the differential user and the distribution coordinate data of the differential user in the user distribution relationship network to be analyzed, may further include the following:
For each differential user, determining each non-differential user belonging to the same user sub-group as the differential user as a candidate related user corresponding to the differential user according to the distribution coordinate data of the differential user in the user distribution relation network to be analyzed;
determining a corresponding user screening rule based on the user sub-population types corresponding to the differential users, wherein the corresponding relationship between the user screening rule and the user sub-population types can be pre-established, for example, the user sub-population type 1 corresponds to the user screening rule 1;
screening candidate related users corresponding to the differential users based on the corresponding user screening rules to form related users corresponding to the differential users; for example, the user sub-population class characterization has a greater degree of tendency to a specific behavior, so that any filtering can be performed among candidate relevant users to form relevant users, or the user sub-population class characterization has a lesser degree of tendency to a specific behavior, so that the user closest to the candidate relevant users (i.e., the closest distribution relationship) can be filtered out as the relevant user.
It will be appreciated that in some embodiments, the user requirement mining method based on big data analysis may further include the following:
extracting a typical user distribution relation network, wherein the typical user distribution relation network can be a user distribution relation network constructed and formed based on typical users;
analyzing original user mining data corresponding to the typical user distribution relation network by using a candidate user mining network, and determining original user mining parameter distribution based on the original user mining data and typical user data characterization vectors of the typical user distribution relation network, wherein the original user mining data is obtained based on typical group distribution characterization vectors corresponding to typical user groups in the typical user distribution relation network, the typical group distribution characterization vectors are used for characterizing the distribution relation of user groups, and the original user mining parameter distribution is used for reflecting distribution coordinate data of differential users of the typical user groups in the typical user distribution relation network;
performing a relationship network adjustment operation on the representative user distribution relationship network to form an adjusted user distribution relationship network (for example, firstly dividing the representative user distribution relationship network to form a plurality of divided sub-areas, then adjusting the distribution coordinate relationship between the divided sub-areas to form an entire user distribution relationship network, in addition, when performing a relationship network adjustment operation, at least one relationship network adjustment operation may be performed to form at least one adjusted user distribution relationship network, so that corresponding subsequent processing may be performed for each adjusted user distribution relationship network), and analyzing adjusted user mining data corresponding to the adjusted user distribution relationship network by using the candidate user mining network, and determining an adjusted user mining parameter distribution based on the adjusted user mining data and an adjusted user data characterization vector of the adjusted user distribution relationship network, wherein the adjusted user mining data is obtained based on an adjusted user population distribution characterization vector corresponding to a representative user population in the adjusted user distribution relationship network, the adjusted user mining parameter distribution is used for characterizing the distribution relationship of the user population, the adjusted user mining parameter distribution is used for reflecting the distribution relationship of the representative user population, and the adjusted user population is based on the typical user distribution relationship of the representative user population, the user distribution relationship is based on the error-related relationship of the user profile data, wherein the typical user distribution relationship is based on the error-related relationship of the user profile data, to perform network analysis;
Analyzing associated mining error indexes corresponding to the candidate user mining network, namely errors before and after adjustment, based on the original user mining parameter distribution and the adjusted user mining parameter distribution;
analyzing target mining error indexes corresponding to the candidate user mining network, namely, the errors between analysis results and actual results, based on the original user mining data, the adjusted user mining data and the differential user identification data corresponding to the typical user distribution relation network;
and performing network optimization operation on the candidate user mining network according to the associated mining error index and the target mining error index to form a target user mining network, wherein the target user mining network is used for analyzing user sub-group category information and distribution coordinate analysis data corresponding to the differential users in the user distribution relation network to be analyzed (e.g. executing the steps S120-S130).
It will be appreciated that, in some embodiments, the steps of using the candidate user mining network to analyze the original user mining data corresponding to the typical user distribution relationship network, and determining the original user mining parameter distribution based on the original user mining data and the typical user data characterization vector of the typical user distribution relationship network may further include the following:
Loading the typical user distribution relation network to load into the candidate user mining network, and mining typical group distribution characterization vectors corresponding to the typical user groups in the typical user distribution relation network by utilizing the candidate user mining network;
analyzing the representative group distribution characterization vector based on a category analysis unit in the candidate user mining network to output original user mining data corresponding to the representative user distribution relation network, namely predicting and outputting the representative group distribution characterization vector;
extracting typical user data characterization vectors which are extracted by a designated filter unit in the candidate user mining network and aim at the typical user distribution relation network, and carrying out multiplication calculation operation on the original user mining data and the typical user data characterization vectors so as to output undetermined user mining parameter distribution corresponding to the typical user distribution relation network, wherein the designated filter unit can be the last filter unit in the candidate user mining network, and the filter unit can comprise one or more filter matrixes or can be a convolution kernel;
Interpolation (upsampling) of the pending user mining parameter distribution is performed to form an original user mining parameter distribution having the same size as the representative user distribution relationship network.
It may be appreciated that, in some embodiments, the step of loading the representative user distribution relationship network to load into the candidate user mining network, and mining the representative population distribution characterization vector corresponding to the representative user population in the representative user distribution relationship network by using the candidate user mining network may further include the following steps:
loading the typical user distribution relation network to load into the candidate user mining network;
digging an overall group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network by utilizing the candidate user mining network;
analyzing the overall user mining data corresponding to the overall group distribution characterization vector by using a category analysis unit in the candidate user mining network;
performing multiplication calculation operation on the whole user mining data and the typical user data characterization vector to output whole user mining parameter distribution corresponding to the typical user distribution relation network, wherein the whole user mining parameter distribution can have information of subgroup types of each typical user;
Dividing the typical user distribution relation network based on the overall user mining parameter distribution to form a first number of typical sub-population relation networks, that is, based on the overall user mining parameter distribution, since the overall user mining parameter distribution has information of sub-population types to which each typical user belongs, the overall user mining parameter distribution can be divided according to corresponding sub-population types, such as dividing typical users corresponding to the same sub-population type together to form one typical sub-population relation network, so that a first number of typical sub-population relation networks can be obtained;
loading the first number of typical sub-population relation networks respectively to load the first number of typical sub-population relation networks into the candidate user mining network, and mining sub-population distribution characterization vectors corresponding to the first number of typical sub-population relation networks respectively by using the candidate user mining network, wherein the sub-population distribution characterization vectors can be used for reflecting the distribution relation of corresponding sub-populations;
and performing aggregation operation on the overall population distribution characterization vector and the sub-population distribution characterization vector corresponding to the first number of typical sub-population relationship networks to form a typical population distribution characterization vector corresponding to the typical user population in the corresponding typical user distribution relationship networks, for example, performing cascade combination operation on the overall population distribution characterization vector and the sub-population distribution characterization vector corresponding to the first number of typical sub-population relationship networks to form a typical population distribution characterization vector, wherein the typical population distribution characterization vector comprises both the sub-population distribution characterization vector of each sub-population of the typical user population and the overall population distribution characterization vector of the typical user population, so that the accuracy of the typical population distribution characterization vector can be enhanced by introducing local learning based on sub-population analysis in a candidate user mining network, and the reliability of the network can be improved.
It will be appreciated that in some embodiments, the candidate user mining network includes a second number of gradient optimization units, each gradient optimization unit including one or more filtering subunits (multiple filtering subunits may be cascade-connected), based on which the step of mining, using the candidate user mining network, the overall population distribution characterization vector corresponding to the typical user population in the typical user distribution relationship network may further include:
determining data to be processed of any one gradient optimizing unit in the second number of gradient optimizing units; in the case that any one gradient optimizing unit is the first gradient optimizing unit, the data to be processed of the any one gradient optimizing unit is the typical user distribution relation network, or may be a relation network mapping feature representation of the typical user distribution relation network, that is, the typical user distribution relation network performs a feature space mapping operation to form a relation network mapping feature representation;
filtering the data to be processed of any one gradient optimizing unit based on one or more filtering subunits in the any one gradient optimizing unit to form an intermediate filtering characterization vector;
Performing aggregation operation, such as cascade combination operation, on the intermediate filtering characterization vector and the data to be processed of any one gradient optimization unit, or alternatively, performing superposition operation to form optimized output data of any one gradient optimization unit, and performing marking operation on the optimized output data of any one gradient optimization unit to mark the optimized output data as the data to be processed of the latter gradient optimization unit;
and marking the optimized output data of the last gradient optimizing unit to be marked as an integral group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network.
For example, the first gradient optimizing unit may perform a filtering operation on the relational network mapping feature representation of the representative user distribution relational network to form a first intermediate filtering characterization vector, the first intermediate filtering characterization vector may perform a superposition operation with the relational network mapping feature representation to form first optimized output data, the second gradient optimizing unit may perform a filtering operation on the first optimized output data to form a second intermediate filtering characterization vector, then the second intermediate filtering characterization vector and the first optimized output data may perform a superposition operation to form second optimized output data, the third gradient optimizing unit may perform a filtering operation on the second optimized output data to form a third intermediate filtering characterization vector, then the third intermediate filtering characterization vector and the second optimized output data may be subjected to a superposition operation to form third optimized output data, and so on, the optimized output data of the last gradient optimizing unit may be obtained, and then may be marked as a corresponding population distribution characterization vector.
It may be appreciated that, in some embodiments, the number of the overall population distribution characterization vectors is a third number, based on which the step of analyzing the overall user mining data corresponding to the overall population distribution characterization vectors by using the category analysis unit in the candidate user mining network may further include the following:
respectively calculating the mean vector parameters corresponding to the third number of the overall population distribution characterization vectors, performing vector construction operation on the mean vector parameters corresponding to the third number of the overall population distribution characterization vectors to form corresponding overall mean characterization vectors, for example, for each of the overall population distribution characterization vectors, performing mean calculation on each vector parameter included in the overall population distribution characterization vector to obtain a corresponding mean vector parameter, and then combining the mean vector parameters corresponding to each of the overall population distribution characterization vectors to form an overall mean characterization vector including the third number of the mean vector parameters, or, in other embodiments, respectively determining the maximum vector parameters corresponding to each of the third number of the overall population distribution characterization vectors, and performing vector construction operation on the maximum vector parameters corresponding to the third number of the overall population distribution characterization vectors to form corresponding overall mean characterization vectors;
Performing excitation mapping operation on the overall mean token vector based on an excitation mapping unit included in the candidate user mining network to form a corresponding excitation mapping feature vector, wherein the excitation mapping unit can include an excitation mapping function, such as an S-type function;
and loading the excitation mapping feature vector to a category analysis unit included in the candidate user mining network, and analyzing the whole user mining data corresponding to the excitation mapping feature vector by using the category analysis unit in the candidate user mining network, namely analyzing and predicting the excitation mapping feature vector by using the category analysis unit.
It may be appreciated that, in some embodiments, the step of performing the network optimization operation on the candidate user mining network according to the associated mining error index and the target mining error index to form a target user mining network may further include the following steps:
calculating a total mining error index corresponding to the candidate user mining network based on the associated mining error index and the target mining error index, for example, weighting and summing calculation can be performed on the associated mining error index and the target mining error index to obtain the total mining error index corresponding to the candidate user mining network;
And optimally adjusting the parameters of the candidate user mining network along the direction of reducing the total mining error index so as to form a target user mining network comprising the optimally adjusted parameters.
It may be appreciated that, in some embodiments, the step of analyzing the associated mining error index corresponding to the candidate user mining network based on the original user mining parameter distribution and the adjusted user mining parameter distribution may further include the following:
performing a parameter distribution adjustment operation corresponding to the relation network adjustment operation on the adjusted user mining parameter distribution (for example, according to a position adjustment corresponding to the relation network adjustment operation, the adjusted user mining parameter distribution may be subjected to a corresponding position adjustment, if in the typical user distribution relation network, the data of the position a and the data of the position B are exchanged, and in the adjusted user mining parameter distribution, the parameters of the position a and the position B may be exchanged, that is, the rotation of the position is implemented), so as to form a corresponding new adjusted user mining parameter distribution;
and performing difference analysis operation on the original user mining parameter distribution and the new adjusted user mining parameter distribution to form an associated mining error index corresponding to the candidate user mining network.
It may be appreciated that, in some embodiments, the step of analyzing the target mining error index corresponding to the candidate user mining network based on the original user mining data, the adjusted user mining data, and the differential user identification data corresponding to the typical user distribution relationship network may further include the following:
calculating original difference information between the original user mining data and the differential user identification data corresponding to the typical user distribution relation network, and analyzing original mining error indexes corresponding to the candidate user mining network based on the original difference information, wherein the original mining error indexes can be positively correlated with the original difference information;
calculating adjustment difference information between the adjustment user mining data and the differential user identification data, and analyzing adjustment mining error indexes of the candidate user mining network based on the adjustment difference information, wherein the adjustment mining error indexes are positively correlated with the adjustment difference information;
and calculating target mining error indexes corresponding to the candidate user mining network based on the original mining error indexes and the adjustment mining error indexes, for example, weighting and summing calculation can be performed on the original mining error indexes and the adjustment mining error indexes so as to obtain the target mining error indexes corresponding to the candidate user mining network.
With reference to fig. 3, the embodiment of the invention further provides a user demand mining device based on big data analysis, which can be applied to the user demand mining system based on big data analysis. Wherein, the user demand mining device based on big data analysis may include:
the user distribution relation network construction module is used for constructing a user distribution relation network to be analyzed based on the user to be analyzed;
the sub-population type data analysis module is used for utilizing a target user mining network to analyze a user population distribution characterization vector corresponding to a user population in the user distribution relation network to be analyzed, and analyzing user sub-population type data corresponding to the user population distribution characterization vector, wherein the user population distribution characterization vector is used for characterizing the distribution relation of the user population, and the user sub-population type data is used for reflecting the user sub-population types corresponding to the user population diversity;
the mining parameter distribution generation module is used for generating a group user mining parameter distribution based on the group user data characterization vector of the user sub-group type data and the user distribution relation network to be analyzed, and the group user mining parameter distribution is used for reflecting distribution coordinate data of the differential users in the user distribution relation network to be analyzed;
The related user determining module is used for determining related users corresponding to the differential users in the user distribution relation network to be analyzed based on the user sub-group types corresponding to the differential users and the distribution coordinate data of the differential users in the user distribution relation network to be analyzed;
the user demand mining module is used for respectively carrying out user demand mining operation based on the user attribute data of each differential user and the user attribute data of the related user of the differential user so as to obtain a user demand mining result corresponding to each differential user, wherein the user demand mining result is used for reflecting the information of interest of the differential user, and the information of interest at least comprises one of text data and image data.
In summary, the user demand mining method and system based on big data analysis provided by the invention can construct a user distribution relation network to be analyzed; analyzing a user group distribution characterization vector corresponding to a user group in a user distribution relation network to be analyzed, and analyzing user sub-group type data corresponding to the user group distribution characterization vector; generating group user mining parameter distribution based on the group user data characterization vector of the user sub-group species data and the user distribution relation network to be analyzed; based on the user sub-group types and the distributed coordinate data, corresponding relevant users are determined; and carrying out user demand mining operation based on the user attribute data of each differential user and the user attribute data of the related user of the differential user respectively to obtain a user demand mining result corresponding to each differential user. Based on the foregoing, in the process of performing the user demand mining operation, not only the user attribute data of the corresponding user is analyzed, but also the corresponding user attribute data of the related user is combined to perform the analysis, so that the analysis basis is more sufficient and more constraint is provided, and therefore, the reliability of the user demand mining can be improved, and the problem of low reliability in the prior art is solved.
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 (9)

1. The user demand mining method based on big data analysis is characterized by comprising the following steps of:
constructing a user distribution relation network to be analyzed based on the users to be analyzed;
the step of constructing the user distribution relation network to be analyzed based on the user to be analyzed comprises the following steps:
determining a plurality of users to be analyzed;
respectively extracting user attribute data of each user to be analyzed, wherein the user attribute data at least comprises static attribute data and dynamic attribute data of the user to be analyzed, the static attribute data at least comprises identity data, the dynamic attribute data at least comprises behavior data, and the user attribute data at least comprises text data;
determining user correlation among users to be analyzed based on the user attribute data, and performing construction operation of a relationship network on the plurality of users to be analyzed based on the user correlation to form a corresponding user distribution relationship network to be analyzed, wherein the distribution relationship among the users to be analyzed is related to the user correlation in the user distribution relationship network to be analyzed;
Analyzing a user group distribution characterization vector corresponding to a user group in the user distribution relation network to be analyzed by utilizing a target user mining network, and analyzing user sub-group type data corresponding to the user group distribution characterization vector, wherein the user group distribution characterization vector is used for characterizing the distribution relation of the user group, and the user sub-group type data is used for reflecting the user sub-group types corresponding to the differential users of the user group;
generating group user mining parameter distribution based on the group user data characterization vector of the user sub-group type data and the user distribution relation network to be analyzed, wherein the group user mining parameter distribution is used for reflecting distribution coordinate data of the differential users in the user distribution relation network to be analyzed;
for each of the differential users, determining relevant users corresponding to the differential users in the user distribution relation network to be analyzed based on the user sub-group types corresponding to the differential users and the distribution coordinate data of the differential users in the user distribution relation network to be analyzed;
and carrying out user demand mining operation based on the user attribute data of each differential user and the user attribute data of the related user of the differential user respectively to obtain a user demand mining result corresponding to each differential user, wherein the user demand mining result is used for reflecting the information of interest of the differential user in the form of text data.
2. The method for mining user requirements based on big data analysis according to claim 1, wherein the step of mining the network by using the target user, analyzing the user population distribution characterization vector corresponding to the user population in the user distribution relation network to be analyzed, and analyzing the user sub-population category data corresponding to the user population distribution characterization vector comprises:
loading the user distribution relation network to be analyzed to the target user mining network, mining an overall group characterization vector corresponding to the user group in the user distribution relation network to be analyzed by using the target user mining network, and analyzing overall group type data corresponding to the overall group characterization vector by using a type analysis unit in the target user mining network, wherein the overall group type data is used for reflecting whether a user belongs to a user group;
extracting group user data characterization vectors which are extracted by a designated filtering unit in the target user mining network and aim at the user distribution relation network to be analyzed, and carrying out multiplication calculation operation on the whole group type data and the group user data characterization vectors so as to output whole group mining parameter distribution corresponding to the user distribution relation network to be analyzed;
Dividing the user distribution relation network to be analyzed based on the overall group mining parameter distribution to form a first number of sub-group distribution relation networks, and mining sub-group representative vectors corresponding to the first number of sub-group distribution relation networks based on the target user mining network;
performing aggregation operation on the overall group characterization vector and subgroup characterization vectors corresponding to the first number of subgroup distribution relation networks to form corresponding user group distribution characterization vectors;
and analyzing the user sub-group type data corresponding to the user group distribution characterization vector.
3. The big data analysis based user demand mining method according to any one of claims 1 to 2, wherein the big data analysis based user demand mining method further comprises:
extracting a typical user distribution relation network;
analyzing original user mining data corresponding to the typical user distribution relation network by using a candidate user mining network, and determining original user mining parameter distribution based on the original user mining data and typical user data characterization vectors of the typical user distribution relation network, wherein the original user mining data is obtained based on typical group distribution characterization vectors corresponding to typical user groups in the typical user distribution relation network, the typical group distribution characterization vectors are used for characterizing the distribution relation of user groups, and the original user mining parameter distribution is used for reflecting distribution coordinate data of differential users of the typical user groups in the typical user distribution relation network;
Performing relation network adjustment operation on the typical user distribution relation network to form an adjustment user distribution relation network, analyzing adjustment user mining data corresponding to the adjustment user distribution relation network by utilizing the candidate user mining network, and determining adjustment user mining parameter distribution based on the adjustment user mining data and adjustment user data characterization vectors of the adjustment user distribution relation network, wherein the adjustment user mining data is obtained based on adjustment group distribution characterization vectors corresponding to typical user groups in the adjustment user distribution relation network, the adjustment group distribution characterization vectors are used for characterizing the distribution relation of user groups, and the adjustment user mining parameter distribution is used for reflecting distribution coordinate data of differential users of the typical user groups in the adjustment user distribution relation network;
analyzing associated mining error indexes corresponding to the candidate user mining network based on the original user mining parameter distribution and the adjusted user mining parameter distribution;
analyzing target mining error indexes corresponding to the candidate user mining network based on the original user mining data, the adjusted user mining data and the differential user identification data corresponding to the typical user distribution relation network;
And carrying out network optimization operation on the candidate user mining network according to the associated mining error index and the target mining error index to form a target user mining network, wherein the target user mining network is used for analyzing user sub-group category information and distribution coordinate analysis data corresponding to the differential users in the user distribution relation network to be analyzed.
4. The method for mining user requirements based on big data analysis according to claim 3, wherein the steps of using candidate user mining networks to analyze original user mining data corresponding to the typical user distribution relation network, and determining original user mining parameter distribution based on the original user mining data and typical user data characterization vectors of the typical user distribution relation network include:
loading the typical user distribution relation network to load into the candidate user mining network, and mining typical group distribution characterization vectors corresponding to the typical user groups in the typical user distribution relation network by utilizing the candidate user mining network;
analyzing the representative group distribution characterization vector based on a category analysis unit in the candidate user mining network so as to output original user mining data corresponding to the representative user distribution relation network;
Extracting typical user data characterization vectors which are extracted by a designated filtering unit in the candidate user mining network and aim at the typical user distribution relation network, and carrying out multiplication calculation operation on the original user mining data and the typical user data characterization vectors so as to output undetermined user mining parameter distribution corresponding to the typical user distribution relation network;
and carrying out interpolation operation on the undetermined user mining parameter distribution to form an original user mining parameter distribution which has the same size as the typical user distribution relation network.
5. The method for mining user requirements based on big data analysis according to claim 4, wherein the step of loading the typical user distribution relation network to load into the candidate user mining network and mining typical population distribution characterization vectors corresponding to the typical user population in the typical user distribution relation network by using the candidate user mining network includes:
loading the typical user distribution relation network to load into the candidate user mining network;
digging an overall group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network by utilizing the candidate user mining network;
Analyzing the overall user mining data corresponding to the overall group distribution characterization vector by using a category analysis unit in the candidate user mining network;
performing multiplication calculation operation on the whole user mining data and the typical user data characterization vector to output whole user mining parameter distribution corresponding to the typical user distribution relation network;
dividing the representative user distribution relationship network based on the overall user mining parameter distribution to form a first number of representative sub-population relationship networks;
loading the first number of typical sub-population relation networks respectively to load the first number of typical sub-population relation networks into the candidate user mining network, and mining sub-population distribution characterization vectors corresponding to the first number of typical sub-population relation networks respectively by utilizing the candidate user mining network;
and performing aggregation operation on the overall population distribution characterization vector and the sub-population distribution characterization vectors corresponding to the first number of typical sub-population relationship networks to form typical population distribution characterization vectors corresponding to the typical user populations in the corresponding typical user distribution relationship networks.
6. The big data analysis based user demand mining method of claim 5, wherein the candidate user mining network comprises a second number of gradient optimization units, each gradient optimization unit comprising one or more filtering subunits;
The step of mining the overall group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network by using the candidate user mining network includes:
determining data to be processed of any one gradient optimizing unit in the second number of gradient optimizing units; in the case that any one gradient optimizing unit is the first gradient optimizing unit, the data to be processed of the any one gradient optimizing unit is the typical user distribution relation network;
filtering the data to be processed of any one gradient optimizing unit based on one or more filtering subunits in the any one gradient optimizing unit to form an intermediate filtering characterization vector;
performing aggregation operation on the intermediate filtering characterization vector and the data to be processed of any one gradient optimization unit to form optimized output data of the any one gradient optimization unit, and performing marking operation on the optimized output data of the any one gradient optimization unit to mark the optimized output data as the data to be processed of the latter gradient optimization unit;
and marking the optimized output data of the last gradient optimizing unit to be marked as an integral group distribution characterization vector corresponding to the typical user group in the typical user distribution relation network.
7. The big data analysis based user demand mining method of claim 5, wherein the number of the overall population distribution characterization vectors is a third number;
the step of analyzing the overall user mining data corresponding to the overall group distribution characterization vector by using a category analysis unit in the candidate user mining network includes:
respectively calculating the average value vector parameters corresponding to the third number of the overall population distribution characterization vectors, and carrying out vector construction operation on the average value vector parameters corresponding to the third number of the overall population distribution characterization vectors to form corresponding overall average value characterization vectors;
performing excitation mapping operation on the whole mean value representation vector based on an excitation mapping unit included in the candidate user mining network so as to form a corresponding excitation mapping feature vector;
and loading the excitation mapping feature vector to load into a category analysis unit included in the candidate user mining network, and analyzing the whole user mining data corresponding to the excitation mapping feature vector by using the category analysis unit in the candidate user mining network.
8. The method for mining user requirements based on big data analysis according to claim 3, wherein the step of performing network optimization operation on the candidate user mining network according to the associated mining error index and the target mining error index to form a target user mining network comprises:
calculating the total mining error index corresponding to the candidate user mining network based on the associated mining error index and the target mining error index;
and optimally adjusting the parameters of the candidate user mining network along the direction of reducing the total mining error index so as to form a target user mining network comprising the optimally adjusted parameters.
9. A user demand mining system based on big data analysis, comprising a processor and a memory, the memory for storing a computer program, the processor for executing the computer program to implement the method of any of claims 1-8.
CN202310404517.6A 2023-04-17 2023-04-17 User demand mining method and system based on big data analysis Active CN116109121B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310404517.6A CN116109121B (en) 2023-04-17 2023-04-17 User demand mining method and system based on big data analysis

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310404517.6A CN116109121B (en) 2023-04-17 2023-04-17 User demand mining method and system based on big data analysis

Publications (2)

Publication Number Publication Date
CN116109121A CN116109121A (en) 2023-05-12
CN116109121B true CN116109121B (en) 2023-06-30

Family

ID=86258350

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310404517.6A Active CN116109121B (en) 2023-04-17 2023-04-17 User demand mining method and system based on big data analysis

Country Status (1)

Country Link
CN (1) CN116109121B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116680323B (en) * 2023-06-20 2024-02-06 深圳市优品投资顾问有限公司 User demand mining method and system based on big data security platform
CN117149846A (en) * 2023-08-16 2023-12-01 湖北中恒电测科技有限公司 Power data analysis method and system based on data fusion
CN117421459B (en) * 2023-12-14 2024-02-23 成都智慧锦城大数据有限公司 Data mining method and system applied to digital city
CN117573742B (en) * 2024-01-15 2024-05-07 浙江华创视讯科技有限公司 User behavior pattern mining method, device and storage medium

Citations (46)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102195899A (en) * 2011-05-30 2011-09-21 中国人民解放军总参谋部第五十四研究所 Method and system for information mining of communication network
CN103269070A (en) * 2013-06-05 2013-08-28 广州中国科学院沈阳自动化研究所分所 Automatic demand response system and automatic demand response method
CN103916293A (en) * 2014-04-15 2014-07-09 浪潮软件股份有限公司 Method for monitoring and analyzing website user behaviors
CN104200275A (en) * 2014-06-24 2014-12-10 国家电网公司 Power utilization mode classification and control method based on user behavior characteristics
CN105468737A (en) * 2015-11-24 2016-04-06 湖北大学 Web service big data analysis method, cloud computing platform and mining system
CN106127503A (en) * 2016-06-06 2016-11-16 广州市邦富软件有限公司 A kind of Analysis of Network Information method based on true social relations and big data
CN106339503A (en) * 2016-09-22 2017-01-18 北京金堤科技有限公司 Interactive display method based on elements and relationships
CN107122864A (en) * 2017-04-28 2017-09-01 国网山东省电力公司泰安供电公司 Power consumer requirement analysis method and device
CN107657491A (en) * 2016-07-25 2018-02-02 长沙有干货网络技术有限公司 A kind of tensor resolution e-commerce user trust recommendation algorithm based on theme
CN108268477A (en) * 2016-12-30 2018-07-10 中国电信股份有限公司 A kind of customer relationship based on ticket determines method and apparatus
CN108281197A (en) * 2018-01-25 2018-07-13 中南大学 A method of relationship between analysis environmental factor and juvenile shortsightedness
CN108320168A (en) * 2017-01-16 2018-07-24 北京金山云网络技术有限公司 A kind of data analysing method and device
CN108448575A (en) * 2018-04-02 2018-08-24 广州供电局有限公司 Power distribution network electric load analysis method and system, computer storage media and equipment
CN108648123A (en) * 2018-07-13 2018-10-12 江苏开放大学（江苏城市职业学院） A method of its management network teaching process of the network teaching platform and utilization based on big data
CN108877774A (en) * 2018-07-12 2018-11-23 四川虹美智能科技有限公司 A kind of data acquisition device, Data Analysis Platform, system and method
CN108964050A (en) * 2018-08-26 2018-12-07 燕山大学 Micro-capacitance sensor dual-layer optimization dispatching method based on Demand Side Response
CN109165996A (en) * 2018-07-18 2019-01-08 浙江大学 Product function feature importance analysis method based on online user's comment
CN109344957A (en) * 2018-08-01 2019-02-15 浙江工业大学 Convolutional neural networks visual analysis method based on comparison in difference
CN109447048A (en) * 2018-12-25 2019-03-08 苏州闪驰数控系统集成有限公司 A kind of artificial intelligence early warning system
CN109583757A (en) * 2018-11-30 2019-04-05 云南民族大学 A kind of points-scoring system of credit system and demand response based on energy internet
CN109885797A (en) * 2019-02-18 2019-06-14 武汉大学 A kind of relational network construction method based on the mapping of more identity spaces
CN109885597A (en) * 2019-01-07 2019-06-14 平安科技（深圳）有限公司 Tenant group processing method, device and electric terminal based on machine learning
CN109902140A (en) * 2019-03-14 2019-06-18 成都小步创想慧联科技有限公司 Event intelligent analysis method, device and user equipment
CN110009216A (en) * 2019-03-28 2019-07-12 阿里巴巴集团控股有限公司 Credit risk subscribers recognition methods and device
CN110019412A (en) * 2017-12-30 2019-07-16 广州明领基因科技有限公司 Distributed mass data digging system based on Agent
CN110096651A (en) * 2019-04-30 2019-08-06 电子科技大学 Visual analysis method based on online social media individual center network
CN110163621A (en) * 2018-02-10 2019-08-23 广州供电局有限公司 A kind of electric power customer service big data DSS
CN110348742A (en) * 2019-07-12 2019-10-18 深圳众赢维融科技有限公司 User data acquisition method, device, electronic equipment and storage medium
CN110350538A (en) * 2019-08-06 2019-10-18 国网上海市电力公司 A kind of micro-grid coordination control method based on active Demand Side Response
CN110442620A (en) * 2019-08-05 2019-11-12 赵玉德 A kind of big data is explored and cognitive approach, device, equipment and computer storage medium
CN110489459A (en) * 2019-08-07 2019-11-22 国网安徽省电力有限公司 A kind of enterprise-level industry number fused data analysis system based on big data platform
CN110555172A (en) * 2019-08-30 2019-12-10 京东数字科技控股有限公司 user relationship mining method and device, electronic equipment and storage medium
CN111061791A (en) * 2019-12-12 2020-04-24 广东食品药品职业学院 Middle and small-sized E-commerce big data platform
CN111159601A (en) * 2019-12-30 2020-05-15 深圳福恋智能信息科技有限公司 Social contact method for visually displaying community based on feature data
CN111324671A (en) * 2020-03-02 2020-06-23 苏州工业园区洛加大先进技术研究院 Biomedical high-speed information processing and analyzing system based on big data technology
CN111683037A (en) * 2020-04-13 2020-09-18 安徽公安职业学院 Intelligent network safety system based on big data analysis
CN111767317A (en) * 2019-04-01 2020-10-13 北京同方软件有限公司 Big data decision analysis method based on data joint service
CN111813834A (en) * 2020-07-14 2020-10-23 滁州职业技术学院 Data mining system and data mining method
CN111833113A (en) * 2020-07-27 2020-10-27 中国平安财产保险股份有限公司 Product recommendation method, device and equipment based on big data and storage medium
CN111860574A (en) * 2020-06-04 2020-10-30 远光软件股份有限公司 Power grid user grouping method and device, storage medium and electronic equipment
CN112115367A (en) * 2020-09-28 2020-12-22 北京百度网讯科技有限公司 Information recommendation method, device, equipment and medium based on converged relationship network
CN115718846A (en) * 2022-12-22 2023-02-28 云南炳暖蔡网络科技有限公司 Big data mining method and system for intelligent interactive network
CN115757556A (en) * 2022-12-05 2023-03-07 徐州思睿晶信息科技有限公司 Big data based network data association analysis method and platform
CN115828160A (en) * 2022-12-20 2023-03-21 徐州思睿晶信息科技有限公司 Data mining method and platform based on big data and cloud computing
CN115905715A (en) * 2022-12-30 2023-04-04 徐州海清信息科技有限公司 Internet data analysis method and platform based on big data and artificial intelligence
CN115964432A (en) * 2023-02-21 2023-04-14 靳佳彬 User operation abnormity analysis method and system based on big data visualization

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007049282A2 (en) * 2005-10-26 2007-05-03 Cortica Ltd. A computing device, a system and a method for parallel processing of data streams
US20140337488A1 (en) * 2013-05-08 2014-11-13 Connectloud Method and Apparatus To Orchestrate Any-vendor IT Infrastructure (Storage) Configuration
US9364151B2 (en) * 2014-03-31 2016-06-14 Elwha Llc Quantified-self machines and circuits reflexively related to food-and-nutrition machines and circuits
US10481195B2 (en) * 2015-12-02 2019-11-19 Machinesense, Llc Distributed IoT based sensor analytics for power line diagnosis
US20160372079A1 (en) * 2015-06-17 2016-12-22 Duk Gwan Ku Personalized color display system through big data analysis and method thereof

Patent Citations (46)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102195899A (en) * 2011-05-30 2011-09-21 中国人民解放军总参谋部第五十四研究所 Method and system for information mining of communication network
CN103269070A (en) * 2013-06-05 2013-08-28 广州中国科学院沈阳自动化研究所分所 Automatic demand response system and automatic demand response method
CN103916293A (en) * 2014-04-15 2014-07-09 浪潮软件股份有限公司 Method for monitoring and analyzing website user behaviors
CN104200275A (en) * 2014-06-24 2014-12-10 国家电网公司 Power utilization mode classification and control method based on user behavior characteristics
CN105468737A (en) * 2015-11-24 2016-04-06 湖北大学 Web service big data analysis method, cloud computing platform and mining system
CN106127503A (en) * 2016-06-06 2016-11-16 广州市邦富软件有限公司 A kind of Analysis of Network Information method based on true social relations and big data
CN107657491A (en) * 2016-07-25 2018-02-02 长沙有干货网络技术有限公司 A kind of tensor resolution e-commerce user trust recommendation algorithm based on theme
CN106339503A (en) * 2016-09-22 2017-01-18 北京金堤科技有限公司 Interactive display method based on elements and relationships
CN108268477A (en) * 2016-12-30 2018-07-10 中国电信股份有限公司 A kind of customer relationship based on ticket determines method and apparatus
CN108320168A (en) * 2017-01-16 2018-07-24 北京金山云网络技术有限公司 A kind of data analysing method and device
CN107122864A (en) * 2017-04-28 2017-09-01 国网山东省电力公司泰安供电公司 Power consumer requirement analysis method and device
CN110019412A (en) * 2017-12-30 2019-07-16 广州明领基因科技有限公司 Distributed mass data digging system based on Agent
CN108281197A (en) * 2018-01-25 2018-07-13 中南大学 A method of relationship between analysis environmental factor and juvenile shortsightedness
CN110163621A (en) * 2018-02-10 2019-08-23 广州供电局有限公司 A kind of electric power customer service big data DSS
CN108448575A (en) * 2018-04-02 2018-08-24 广州供电局有限公司 Power distribution network electric load analysis method and system, computer storage media and equipment
CN108877774A (en) * 2018-07-12 2018-11-23 四川虹美智能科技有限公司 A kind of data acquisition device, Data Analysis Platform, system and method
CN108648123A (en) * 2018-07-13 2018-10-12 江苏开放大学（江苏城市职业学院） A method of its management network teaching process of the network teaching platform and utilization based on big data
CN109165996A (en) * 2018-07-18 2019-01-08 浙江大学 Product function feature importance analysis method based on online user's comment
CN109344957A (en) * 2018-08-01 2019-02-15 浙江工业大学 Convolutional neural networks visual analysis method based on comparison in difference
CN108964050A (en) * 2018-08-26 2018-12-07 燕山大学 Micro-capacitance sensor dual-layer optimization dispatching method based on Demand Side Response
CN109583757A (en) * 2018-11-30 2019-04-05 云南民族大学 A kind of points-scoring system of credit system and demand response based on energy internet
CN109447048A (en) * 2018-12-25 2019-03-08 苏州闪驰数控系统集成有限公司 A kind of artificial intelligence early warning system
CN109885597A (en) * 2019-01-07 2019-06-14 平安科技（深圳）有限公司 Tenant group processing method, device and electric terminal based on machine learning
CN109885797A (en) * 2019-02-18 2019-06-14 武汉大学 A kind of relational network construction method based on the mapping of more identity spaces
CN109902140A (en) * 2019-03-14 2019-06-18 成都小步创想慧联科技有限公司 Event intelligent analysis method, device and user equipment
CN110009216A (en) * 2019-03-28 2019-07-12 阿里巴巴集团控股有限公司 Credit risk subscribers recognition methods and device
CN111767317A (en) * 2019-04-01 2020-10-13 北京同方软件有限公司 Big data decision analysis method based on data joint service
CN110096651A (en) * 2019-04-30 2019-08-06 电子科技大学 Visual analysis method based on online social media individual center network
CN110348742A (en) * 2019-07-12 2019-10-18 深圳众赢维融科技有限公司 User data acquisition method, device, electronic equipment and storage medium
CN110442620A (en) * 2019-08-05 2019-11-12 赵玉德 A kind of big data is explored and cognitive approach, device, equipment and computer storage medium
CN110350538A (en) * 2019-08-06 2019-10-18 国网上海市电力公司 A kind of micro-grid coordination control method based on active Demand Side Response
CN110489459A (en) * 2019-08-07 2019-11-22 国网安徽省电力有限公司 A kind of enterprise-level industry number fused data analysis system based on big data platform
CN110555172A (en) * 2019-08-30 2019-12-10 京东数字科技控股有限公司 user relationship mining method and device, electronic equipment and storage medium
CN111061791A (en) * 2019-12-12 2020-04-24 广东食品药品职业学院 Middle and small-sized E-commerce big data platform
CN111159601A (en) * 2019-12-30 2020-05-15 深圳福恋智能信息科技有限公司 Social contact method for visually displaying community based on feature data
CN111324671A (en) * 2020-03-02 2020-06-23 苏州工业园区洛加大先进技术研究院 Biomedical high-speed information processing and analyzing system based on big data technology
CN111683037A (en) * 2020-04-13 2020-09-18 安徽公安职业学院 Intelligent network safety system based on big data analysis
CN111860574A (en) * 2020-06-04 2020-10-30 远光软件股份有限公司 Power grid user grouping method and device, storage medium and electronic equipment
CN111813834A (en) * 2020-07-14 2020-10-23 滁州职业技术学院 Data mining system and data mining method
CN111833113A (en) * 2020-07-27 2020-10-27 中国平安财产保险股份有限公司 Product recommendation method, device and equipment based on big data and storage medium
CN112115367A (en) * 2020-09-28 2020-12-22 北京百度网讯科技有限公司 Information recommendation method, device, equipment and medium based on converged relationship network
CN115757556A (en) * 2022-12-05 2023-03-07 徐州思睿晶信息科技有限公司 Big data based network data association analysis method and platform
CN115828160A (en) * 2022-12-20 2023-03-21 徐州思睿晶信息科技有限公司 Data mining method and platform based on big data and cloud computing
CN115718846A (en) * 2022-12-22 2023-02-28 云南炳暖蔡网络科技有限公司 Big data mining method and system for intelligent interactive network
CN115905715A (en) * 2022-12-30 2023-04-04 徐州海清信息科技有限公司 Internet data analysis method and platform based on big data and artificial intelligence
CN115964432A (en) * 2023-02-21 2023-04-14 靳佳彬 User operation abnormity analysis method and system based on big data visualization

Also Published As

Publication number Publication date
CN116109121A (en) 2023-05-12

Similar Documents

Publication Publication Date Title
CN116109121B (en) User demand mining method and system based on big data analysis
Cai et al. An under‐sampled software defect prediction method based on hybrid multi‐objective cuckoo search
Yang et al. A unified semi-supervised community detection framework using latent space graph regularization
US9536201B2 (en) Identifying associations in data and performing data analysis using a normalized highest mutual information score
CN116126945B (en) Sensor running state analysis method and system based on data analysis
CN113673697A (en) Model pruning method and device based on adjacent convolution and storage medium
Gasimov et al. Separation via polyhedral conic functions
CN116109630B (en) Image analysis method and system based on sensor acquisition and artificial intelligence
CN111090807A (en) Knowledge graph-based user identification method and device
CN115795000A (en) Joint similarity algorithm comparison-based enclosure identification method and device
Silva et al. An instance selection method for large datasets based on markov geometric diffusion
Nawalaniec Classifying and analysis of random composites using structural sums feature vector
CN115828160A (en) Data mining method and platform based on big data and cloud computing
Dommaraju et al. Identifying topological prototypes using deep point cloud autoencoder networks
Li et al. An ensemble clustering framework based on hierarchical clustering ensemble selection and clusters clustering
WO2016093839A1 (en) Structuring of semi-structured log messages
Lee et al. Inverse resolution limit of partition density and detecting overlapping communities by link-surprise
EP4227855A1 (en) Graph explainable artificial intelligence correlation
CN108830302B (en) Image classification method, training method, classification prediction method and related device
Maheswari et al. Kernelized Spectral Clustering based Conditional MapReduce function with big data
Fabian et al. Estimating the execution time of the coupled stage in multiscale numerical simulations
CN114529001A (en) Index evaluation method and device of quantum algorithm, terminal and storage medium
CN112750047A (en) Behavior relation information extraction method and device, storage medium and electronic equipment
Beavers et al. Data Nuggets: A Method for Reducing Big Data While Preserving Data Structure
CN116738396B (en) Artificial intelligence-based landmark quasi document input method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant