CN117333190A - Vendor screening and evaluating method based on knowledge graph and neural network - Google Patents

Vendor screening and evaluating method based on knowledge graph and neural network Download PDF

Info

Publication number
CN117333190A
CN117333190A CN202311187387.1A CN202311187387A CN117333190A CN 117333190 A CN117333190 A CN 117333190A CN 202311187387 A CN202311187387 A CN 202311187387A CN 117333190 A CN117333190 A CN 117333190A
Authority
CN
China
Prior art keywords
knowledge graph
data
knowledge
model
neural network
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202311187387.1A
Other languages
Chinese (zh)
Inventor
卢晓凯
封军
王汝舵
姚丽
高俊
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.)
Anhui High Quality Mining Technology Development Co ltd
Original Assignee
Anhui High Quality Mining Technology Development Co ltd
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 Anhui High Quality Mining Technology Development Co ltd filed Critical Anhui High Quality Mining Technology Development Co ltd
Priority to CN202311187387.1A priority Critical patent/CN117333190A/en
Publication of CN117333190A publication Critical patent/CN117333190A/en
Pending legal-status Critical Current

Links

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
    • G06Q30/00—Commerce
    • G06Q30/01—Customer relationship services
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G06N5/022—Knowledge engineering; Knowledge acquisition

Abstract

The invention is suitable for the technical field of data processing, and provides a supplier screening evaluation method based on a knowledge graph and a neural network, which comprises the following steps: constructing a knowledge graph model oriented to the project, and constructing a provider knowledge graph by using provider data; the method comprises the steps of utilizing item and user tag information, capturing low-order and high-order features through knowledge graph representation learning, embedding semantic information of entities and relations in two knowledge graphs into a low-dimensional vector space, and obtaining unified representation of the item and the user; extracting potential features of the items and the users by using a deep neural network and a recurrent neural network added with an attention mechanism; and predicting corresponding scores by utilizing the learned potential characteristics of the users and the items, sorting according to the score results, and matching and recommending the suppliers and the users. According to the invention, the knowledge graph database is established, the data is vectorized, the provider screening data model is constructed, the high-quality providers are screened and ranked, and the bid-seeking purchasing transaction is facilitated.

Description

Vendor screening and evaluating method based on knowledge graph and neural network
Technical Field
The invention relates to the technical field of data processing, in particular to a vendor screening and evaluating method based on a knowledge graph and a neural network.
Background
The chinese patent with application No. 202010407432.X discloses a computer screening device, method and storage medium for screening suppliers, comprising an acquisition module for acquiring characteristic parameters of reactant stream aging, article out-of-stock conditions, article quality conditions; the determining module is used for determining the total score corresponding to the characteristic parameter and the backorder rate score corresponding to the supplier backorder rate according to the pre-stored corresponding relation; a selection module for selecting the smaller of the overall score and the backorder score as a final score for the supplier; and the first judging module is used for judging whether the final score meets the supply requirement. By acquiring the characteristic parameters for reflecting the logistics ageing, the goods shortage condition and the goods quality condition of the suppliers, the characteristic parameters can directly influence the goods supply condition of the suppliers, further influence the experience of users, and are more suitable for the non-stock mode. The lack of compliance of the supplier with the needs of the buyer in the bidding purchase is detrimental to facilitating the bidding purchase transaction of the supplier with the buyer. Therefore, there is a need to provide a vendor screening evaluation method based on knowledge maps and neural networks, which aims to solve the above problems.
Disclosure of Invention
Aiming at the defects existing in the prior art, the invention aims to provide a supplier screening and evaluating method based on a knowledge graph and a neural network, so as to solve the problems existing in the background art.
The invention is realized in such a way that a provider screening evaluation method based on a knowledge graph and a neural network comprises the following steps:
constructing a project-oriented knowledge graph model by preprocessing data, defining a model, importing data and visualizing the data, and constructing a provider knowledge graph by using provider data; the method comprises the steps that features of project entities are needed to be reflected through project embedding, a knowledge graph centered on the project is constructed firstly, and then KGCN is utilized to perform representation learning on the knowledge graph, so that project entity representation vectors are obtained;
the method comprises the steps of utilizing item and user tag information, capturing low-order and high-order features through knowledge graph representation learning, and embedding semantic information of entities and relations in two knowledge graphs into a low-dimensional vector space, so that unified representation of the item and the user is obtained;
extracting potential features of the items and the users by using a deep neural network and a recurrent neural network added with an attention mechanism; and predicting corresponding scores by utilizing the learned potential characteristics of the users and the items, sorting according to the score results, and matching and recommending the suppliers and the users.
As a further scheme of the invention: when the data is preprocessed, the data is required to be arranged, cleaned and formatted so as to ensure the quality and consistency of the data.
As a further scheme of the invention: when the model is defined, the construction mode of the knowledge graph model needs to be determined, and the structure and the semantics of the model are defined.
As a further scheme of the invention: when the data is visualized, a knowledge graph model is presented through a graph visualization tool, so that a user can conveniently inquire and analyze the data; then, the accuracy and the integrity of the knowledge graph model need to be evaluated to ensure that the model can meet the requirements of purchasing business.
As a further scheme of the invention: the KGCN is a graph annotation meaning network applied to the knowledge graph, is used for capturing topological structure and entity information in the knowledge graph, and is used for gathering information of neighbor nodes of the knowledge graph with prejudice and integrating the information into the characteristics of a given entity in the knowledge graph when calculating the characteristics of the entity, the model is of a multi-layer structure, low-order characteristics of the entity can be obtained when the model is of a low-level structure, and high-order information of the entity can be mined when the model is of a high-level structure.
As a further scheme of the invention: in predicting the score, the inner product of the user potential vector u (i) and the item potential vector v (j) is equal to r (ij) using a matrix decomposition method: r (ij) =u (i) ·v (j), the inner product of two potential feature vectors learned from the mixed deep structure is used to obtain a user's scoring prediction for the item.
As a further scheme of the invention: the method further comprises the steps of: and collecting feedback information, and improving the accuracy and efficiency of the recommendation system through the feedback information of the user.
Compared with the prior art, the invention has the beneficial effects that:
according to the method, the knowledge graph model oriented to the project is constructed, and the knowledge graph of the provider is constructed by using provider data; the method comprises the steps of utilizing item and user tag information, capturing low-order and high-order features through knowledge graph representation learning, and embedding semantic information of entities and relations in two knowledge graphs into a low-dimensional vector space, so that unified representation of the item and the user is obtained; extracting potential features of the items and the users by using a deep neural network and a recurrent neural network added with an attention mechanism; and predicting corresponding scores by utilizing the learned potential characteristics of the users and the items, sorting according to the score results, matching and recommending the suppliers and the users, and screening out high-quality suppliers, thereby being beneficial to facilitating bidding purchase transaction.
Drawings
FIG. 1 is a flow chart of a vendor screening evaluation method based on knowledge maps and neural networks.
FIG. 2 is a graph of relationships between systems in a vendor screening evaluation method based on knowledge maps and neural networks.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more clear, the present invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Specific implementations of the invention are described in detail below in connection with specific embodiments.
As shown in fig. 1 and 2, the embodiment of the invention provides a vendor screening evaluation method based on a knowledge graph and a neural network, which comprises the following steps:
s100, constructing a project-oriented knowledge graph model by preprocessing data, defining a model, importing data and visualizing the data, and constructing a provider knowledge graph by using provider data; the method comprises the steps that features of project entities are needed to be reflected through project embedding, a knowledge graph centered on the project is constructed firstly, and then KGCN is utilized to perform representation learning on the knowledge graph, so that project entity representation vectors are obtained;
s200, utilizing item and user tag information, capturing low-order and high-order features through knowledge graph representation learning, and embedding semantic information of entities and relations in the two knowledge graphs into a low-dimensional vector space so as to obtain unified representation of the item and the user;
s300, extracting potential characteristics of items and users by using a deep neural network and a recurrent neural network added with an attention mechanism; and predicting corresponding scores by utilizing the learned potential characteristics of the users and the items, sorting according to the score results, and matching and recommending the suppliers and the users.
In the embodiment of the invention, a Knowledge Graph (knowledgegraph) is a semantic network, nodes of the semantic network represent entities, edges of the semantic network represent various semantic relations among the entities, and a data module system of the Graph is established, wherein the data system comprises: basic information system, operation information system, purchasing transaction system, enterprise risk system and forecast information system. And establishing a knowledge graph database, constructing a provider screening data model through data vectorization, and finally screening a certain number of high-quality providers according to the model and ranking.
In the embodiment of the invention, when the data is preprocessed, the data is required to be processed by arrangement, cleaning and formatting so as to ensure the quality and consistency of the data. When the model is defined, the construction mode of the knowledge graph model, such as entity, relationship, attribute and the like, needs to be determined, and the structure and the semantics of the model are defined. Then, the processed data are imported into a knowledge graph model so as to realize data visualization, and the knowledge graph model is presented through a graph visualization tool when the data are visualized, so that a user can conveniently inquire and analyze the data; then, the accuracy and the integrity of the knowledge graph model need to be evaluated to ensure that the model can meet the requirements of purchasing business. And finally, establishing a knowledge graph of the platform data: constructing a knowledge graph of the provider by using the existing data such as the purchasing company of the provider, including the product and service information of the provider and the business entity relationship of the provider, including establishing a knowledge graph information system index and constructing a knowledge graph ontology relationship model;
in the embodiment of the invention, the KGCN is used for carrying out representation learning on the knowledge graph to obtain the project entity representation vector, the KGCN is a graph annotation meaning network applied to the knowledge graph and is used for capturing the topological structure and entity information in the knowledge graph, when the characteristics of a given entity in the knowledge graph are calculated, the information of neighbor nodes of the knowledge graph are combined together with prejudice and are integrated into the characteristics of the entity, the model is of a multi-layer structure, the low-level characteristics of the entity can be obtained in the low-level process, and the high-level information of the entity can be mined in the high-level process.
In the embodiment of the invention, when predicting the score, a matrix decomposition method is needed, and the inner product of the user potential vector u (i) and the item potential vector v (j) is equal to r (ij): r (ij) =u (i) ·v (j), the inner product of two potential feature vectors learned from the mixed deep structure is used to obtain the scoring prediction of the user on the project, and finally the scoring result is ranked, the provider is matched with the user, and then the recommendation is performed through the short message. And finally, feedback information is acquired, and the accuracy and efficiency of the recommendation system are improved through the feedback information of the user.
The foregoing description of the preferred embodiments of the present invention should not be taken as limiting the invention, but rather should be understood to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
It should be understood that, although the steps in the flowcharts of the embodiments of the present invention are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in various embodiments may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor do the order in which the sub-steps or stages are performed necessarily performed in sequence, but may be performed alternately or alternately with at least a portion of the sub-steps or stages of other steps or other steps.
Those skilled in the art will appreciate that all or part of the processes in the methods of the above embodiments may be implemented by a computer program for instructing relevant hardware, where the program may be stored in a non-volatile computer readable storage medium, and where the program, when executed, may include processes in the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
Other embodiments of the present disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure. This application is intended to cover any adaptations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

Claims (7)

1. The provider screening and evaluating method based on the knowledge graph and the neural network is characterized by comprising the following steps of:
constructing a project-oriented knowledge graph model by preprocessing data, defining a model, importing data and visualizing the data, and constructing a provider knowledge graph by using provider data; the method comprises the steps that features of project entities are needed to be reflected through project embedding, a knowledge graph centered on the project is constructed firstly, and then KGCN is utilized to perform representation learning on the knowledge graph, so that project entity representation vectors are obtained;
the method comprises the steps of utilizing item and user tag information, capturing low-order and high-order features through knowledge graph representation learning, and embedding semantic information of entities and relations in two knowledge graphs into a low-dimensional vector space, so that unified representation of the item and the user is obtained;
extracting potential features of the items and the users by using a deep neural network and a recurrent neural network added with an attention mechanism; and predicting corresponding scores by utilizing the learned potential characteristics of the users and the items, sorting according to the score results, and matching and recommending the suppliers and the users.
2. The knowledge-graph and neural-network-based vendor screening and evaluation method according to claim 1, wherein the preprocessing of the data requires the data to be sorted, cleaned and formatted to ensure the quality and consistency of the data.
3. The method for evaluating the selection of suppliers based on the knowledge graph and the neural network according to claim 1, wherein the knowledge graph model is constructed in a manner required to be determined and the structure and the semantics of the model are defined when the model is defined.
4. The provider screening evaluation method based on the knowledge graph and the neural network according to claim 1, wherein the knowledge graph model is presented through a graph visualization tool when the data is visualized, so that a user can conveniently inquire and analyze the data; then, the accuracy and the integrity of the knowledge graph model need to be evaluated to ensure that the model can meet the requirements of purchasing business.
5. The provider screening evaluation method based on a knowledge graph and a neural network according to claim 1, wherein the KGCN is a graph annotation meaning network applied to the knowledge graph, and is used for capturing topological structure and entity information in the knowledge graph, and when calculating the characteristics of a given entity in the knowledge graph, the information of neighboring nodes is collected with bias and integrated into the characteristics of the entity, the model is a multi-layer structure, the low-level characteristics of the entity can be obtained at the low level, and the high-level information of the entity can be mined at the high level.
6. The knowledge-graph and neural-network-based provider screening evaluation method according to claim 1, wherein the inner product of the user potential vector u (i) and the item potential vector v (j) is equal to r (ij) by a matrix decomposition method when predicting the score: r (ij) =u (i) ·v (j), the inner product of two potential feature vectors learned from the mixed deep structure is used to obtain a user's scoring prediction for the item.
7. The knowledge-graph and neural-network-based vendor screening evaluation method of claim 1, further comprising: and collecting feedback information, and improving the accuracy and efficiency of the recommendation system through the feedback information of the user.
CN202311187387.1A 2023-09-14 2023-09-14 Vendor screening and evaluating method based on knowledge graph and neural network Pending CN117333190A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311187387.1A CN117333190A (en) 2023-09-14 2023-09-14 Vendor screening and evaluating method based on knowledge graph and neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311187387.1A CN117333190A (en) 2023-09-14 2023-09-14 Vendor screening and evaluating method based on knowledge graph and neural network

Publications (1)

Publication Number Publication Date
CN117333190A true CN117333190A (en) 2024-01-02

Family

ID=89274609

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311187387.1A Pending CN117333190A (en) 2023-09-14 2023-09-14 Vendor screening and evaluating method based on knowledge graph and neural network

Country Status (1)

Country Link
CN (1) CN117333190A (en)

Similar Documents

Publication Publication Date Title
Han et al. Artificial intelligence in business-to-business marketing: a bibliometric analysis of current research status, development and future directions
CN101918961B (en) Evaluating public records of supply transactions
Even et al. Utility-driven assessment of data quality
US20120078678A1 (en) Method and system for estimation and analysis of operational parameters in workflow processes
CN102906686A (en) Evaluating public records of supply transactions for financial investment decisions
Chirra et al. Evaluation of supply chain flexibility in automobile industry with fuzzy DEMATEL approach
Jahani et al. Data science and big data analytics: A systematic review of methodologies used in the supply chain and logistics research
Akerkar Advanced data analytics for business
Mamoudan et al. Factor identification for insurance pricing mechanism using data mining and multi criteria decision making
Wu et al. The state of lead scoring models and their impact on sales performance
Hasheminejad et al. Data mining techniques for analyzing bank customers: A survey
CN117333190A (en) Vendor screening and evaluating method based on knowledge graph and neural network
Mehlstäubl et al. Data mining in product portfolio and variety management–literature review on use cases and research potentials
Keskin A novel perspective to sustainable supplier performance evaluation problem: A case study based on determining the optimum number of clusters
Sharawi et al. Utilization of data visualization for knowledge discovery in modern logistic service companies
Mutha et al. Determining assortments of used products for B2B transactions in reverse supply chain
Granov Customer loyalty, return and churn prediction through machine learning methods: for a Swedish fashion and e-commerce company
CN113297517A (en) Click rate estimation and model training method, system and device
Altuntas et al. A novel technology intelligence tool based on utility mining
CN111415213A (en) Cognitive purchasing
Suh Exploring the Impact of Data Quality on Business Performance in CRM Systems for Home Appliance Business
JP2005135167A (en) Device, method, and program for data analysis
Praveen et al. Big Mart Sales using Hybrid Learning Framework with Data Analysis
Sharma Identifying Factors Contributing to Lead Conversion Using Machine Learning to Gain Business Insights
Nguyen Exploring applications of big data analytics in supply chain management

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