CN111612549A - Construction method of platform operation service system - Google Patents

Construction method of platform operation service system Download PDF

Info

Publication number
CN111612549A
CN111612549A CN202010471261.7A CN202010471261A CN111612549A CN 111612549 A CN111612549 A CN 111612549A CN 202010471261 A CN202010471261 A CN 202010471261A CN 111612549 A CN111612549 A CN 111612549A
Authority
CN
China
Prior art keywords
enterprise
portrait
service provider
service
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.)
Granted
Application number
CN202010471261.7A
Other languages
Chinese (zh)
Other versions
CN111612549B (en
Inventor
任智军
范凯波
樊辉
高宇强
黄静
朱益新
王洁
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.)
Hantang Xintong Beijing Consulting Co ltd
Cqc Intime Testing Technology Co ltd
Original Assignee
Hantang Xintong Beijing Consulting Co ltd
Cqc Intime Testing Technology 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 Hantang Xintong Beijing Consulting Co ltd, Cqc Intime Testing Technology Co ltd filed Critical Hantang Xintong Beijing Consulting Co ltd
Priority to CN202010471261.7A priority Critical patent/CN111612549B/en
Publication of CN111612549A publication Critical patent/CN111612549A/en
Application granted granted Critical
Publication of CN111612549B publication Critical patent/CN111612549B/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
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/25—Integrating or interfacing systems involving database management systems
    • G06F16/254—Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • 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
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • 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
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Educational Administration (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a method for constructing a platform operation service system, which comprises the following steps: collecting data to construct a basic database; constructing a service provider management system; establishing an introduced customer flow pool to form a transaction database; a provider mining base engine; constructing a business mining system by utilizing the business mining basic engine, the basic database and the transaction database, and intelligently matching the business mining system with the service provider management system; potential business opportunities of enterprise users are mined by the business opportunity mining system, and the potential business opportunities are automatically pushed to the matched service provider; and enabling the matched service provider to provide services for the corresponding selected enterprise users according to the potential business opportunities. The method of the present invention can make the platform operation service system integrate the platform service input end and the platform service output end, and realize the benign operation and the deepened application of the platform.

Description

Construction method of platform operation service system
Technical Field
The invention relates to a platform operation service system, in particular to a construction method of the platform operation service system, and belongs to the technical field of computers.
Background
With the rapid development of science and technology and the internet, a platform for the services of small and medium-sized micro enterprises is gradually raised, however, on the basis of building a platform system architecture, the key for exerting the maximum value of the platform lies in building a matched platform operation system, and by developing a series of operation services, the user experience is improved, the platform is continuously energized, and the platform operation target and the like are finally completed. The platform construction and operation are only the beginning, and the popularization, application and deepening practicality are the starting point and the fundamental power of the platform construction, so that how to build an operation service system with high market applicability and high conversion rate and strengthen the popularization and application of the platform becomes a key problem. In order to improve the quality of the operation service, generally, the platform operation system mainly focuses on the user operation and the activity operation, and basically takes a series of measures oriented to improve the service conversion rate as a value.
The existing platform operation service system puts emphasis on user management, activity marketing and other aspects, the operation service system measures the value of business services according to the algorithm logic of 'flow rate conversion rate passenger unit price', and the direct consequence of carrying out operation service for the value judgment standard is that enterprises only pay attention to the simplified conversion rate but miss huge benefits brought by service repeated purchase and full life cycle service, and meanwhile, the viscosity of customers and the deep application of the platform are greatly influenced. Secondly, due to the influence of the guidance of the operation value, a platform operation system lacks of integral planning, mostly only service objects are considered, but the service providers lack of energy and management, and the delivery capacity, credit evaluation and the like of the service providers directly influence the application of the platform to a certain extent. Finally, the traditional platform operation system has relatively little concern on the data asset value, and image analysis, intelligent service recommendation and the like cannot be performed on the service enterprises in the operation process, so that the popularization and application of the platform are restricted to a certain extent.
Disclosure of Invention
The invention mainly aims to provide a method for constructing a platform operation service system, thereby overcoming the defects in the prior art.
In order to achieve the above purpose, the invention provides the following technical scheme:
the embodiment of the invention provides a method for constructing a platform operation service system, which comprises the following steps:
collecting data to construct a basic database, wherein the collected data comprises any one or combination of more of industrial and commercial data, license data, intellectual property data, public opinion data and policy data;
constructing a service provider management system, which comprises a service provider admission auditing module, a service provider imaging module, a service provider retrieving module, a service provider credit evaluation module and a service provider client;
establishing an introduced customer flow pool, forming a transaction database, and then establishing a product portrait and an enterprise portrait of an enterprise user;
providing a business mining basic engine, wherein the business mining basic engine is a multi-source fusion repeated purchasing business engine based on an association rule algorithm and a similarity algorithm;
constructing a business mining system by utilizing the business mining basic engine, the basic database and the transaction database, and intelligently matching the business mining system with the service provider management system;
potential business opportunities of enterprise users are mined by the business opportunity mining system, the potential business opportunities are automatically pushed to the matched service providers, and the business opportunity mining system is optimized according to business opportunity effectiveness evaluation fed back by the matched service providers;
and enabling the matched service provider to provide services for the corresponding selected enterprise users according to the potential business opportunities, and enabling the service provider management system to optimize according to the service evaluation fed back by the selected enterprise users.
In some embodiments, the method of construction comprises: obtaining various types of structured, semi-structured and unstructured mass data at least based on social network data and mobile internet data, then placing the mass data into Hive through an ETL tool, processing and processing the data by using Mapreduce and Spark, and constructing a platform big data base, namely obtaining the basic database.
The service provider admission auditing module is used for realizing warehousing and management of service providers by setting service provider admission standards and index thresholds, wherein the service provider admission standards and index thresholds comprise service provider establishment years, registered capital, enterprise scale, whether more service disputes exist or not and whether the service providers enter a loss-of-trust blacklist or not;
the service provider portrait module is at least used for portrait processing from basic information, delivery capacity and service level of a service provider;
the service provider retrieval module is at least used for collecting service providers according to the processing result of the service provider image module and the service provider characteristics, wherein the service provider characteristics comprise that an enterprise achieves a set scale and has no bad records;
the service provider credit evaluation module is used for performing credit evaluation according to the processing result of the service provider image module and a credit evaluation index system based on AHP and a fuzzy comprehensive evaluation method, wherein the credit evaluation index system comprises innovation development capability status, service level, credit history, performance capability, financial status and management analysis;
the service provider client is at least used for providing a port for auditing, information submitting and feedback for a service provider;
the process of credit evaluation based on the AHP and the fuzzy comprehensive evaluation method comprises the following steps:
A. constructing a relevant judgment matrix of a credit evaluation index system, and constructing a judgment matrix X ═ Xij](i, j ═ 1, 2, 3.., n), where X isij＝Xi/XjMeans XiRelative to XjThe importance of the judgment matrix is sequentially constructed for each level;
Figure BDA0002513073360000031
B. according to the following formulas 1) to 4), determining relative weight coefficients of a target layer service provider credit evaluation, a secondary index layer and a specific measurement index layer, wherein the secondary index layer comprises innovation development capability status, service level, credit history, performance capability, financial status and management analysis, and the specific measurement index layer comprises research and development invested number, liquidity and management team number;
XW＝λmaxw type (1)
Figure BDA0002513073360000032
Figure BDA0002513073360000033
W＝(w1，w2，…，wn)TFormula (4)
X in the formula (1) is a judgment matrix and lambdamaxTo determine the maximum eigenvalue of the matrix, x in equation (2)ijIs an index xiRelative to xjOf importance, w in formula (3)iW is a weight vector in the formula (1) and the formula (4), and W is a weight vector in the formula (2) and the formula (3)
Figure BDA0002513073360000035
The weight of the index layer relative to the second-level index layer is measured specifically;
C. carrying out consistency check on the calculation model according to the following formula 5), calculating a model consistency index CI, determining a corresponding random consistency index RI, calculating a consistency ratio CR, and continuously correcting the model until the CR is less than 0.1, wherein:
Figure BDA0002513073360000034
wherein λ ismaxTo determine the maximum eigenvalue of the matrix, n is the number of calculation indices.
In some embodiments, the method of construction comprises: introducing a large amount of customer flow into enterprise users at least through multi-channel API docking and promotion activities to form an introduced customer flow pool, judging through introduced strategy rules to form a transaction database, and then cleaning and processing data to construct product figures and enterprise figures of the enterprise users;
the introduction strategy is at least used for screening enterprise users through channel classification marking, abnormal operation and risk enterprise elimination and service history blacklist enterprise to establish enterprise ID and enrich a transaction database;
the channel classification mark is used for establishing source and preference basic data of enterprise users so as to help introduce and manage a transaction database;
the enterprise with the removed operation abnormity and the risk is used for improving the safety and the effectiveness of the transaction database by filtering enterprise users with higher risk or operation abnormity;
the enterprise with the service history rejection blacklist is used for rejecting the enterprise with bad service history;
the enterprise ID is a unique identification code of an enterprise user, and is at least established by comparing, matching and text processing enterprise names in business data of different channel sources in the flow pool with a business name library.
In some embodiments, the method of construction comprises: based on the transaction database and the portrait processing label system, a multi-dimensional portrait is constructed through a rule engine and a machine learning algorithm, an enterprise portrait and a product portrait model of an enterprise user are optimized by combining expert feedback and a weight model, and then the product portrait and the enterprise portrait of the enterprise user are obtained by utilizing the optimized enterprise portrait and the optimized product portrait model.
In some embodiments, the method of construction comprises:
s1: establishing an image processing label system, wherein the image processing label system comprises a basic label, an operation label, a product label and a risk label;
s2: forming an enterprise portrait and a product portrait feature word bank based on basic data, wherein the basic data comprises industrial and commercial data and intellectual property data of enterprise users;
s3: selecting an enterprise sample, and matching and processing the enterprise portrait based on a regular expression according to the portrait processing rule base;
s4: aiming at a transaction standardized product, establishing a product portrait processing rule, wherein the product portrait processing rule comprises that the product is suitable for the industry, suitable regions, scale and age;
s5: and selecting enterprise IDs in the same industry and the portrait processing result in the step S3, obtaining similar enterprises based on a K-Means clustering algorithm, and further optimizing the portrait processing rule.
In some embodiments, the method of construction comprises:
I. acquiring a related product rule and a corresponding probability from historical transaction data based on an FP-tree algorithm;
II. Finding out the associated products of the purchased products according to the association rules;
III, calculating the similarity of the product portrait and the enterprise portrait, sorting according to the similarity calculation result, and finding a recommended product according to a set threshold;
and IV, filtering the products by using a filter, and combining the enterprises and the products to form a recommender library.
In some embodiments, the step I comprises:
1) selecting an enterprise sample, and acquiring all historical transaction records based on enterprise IDs;
2) establishing a standardized service product name library, standardizing the purchased products of the sample enterprise based on standardized product names, forming an original product combination according to transaction time, and generating a training data set;
3) setting a support threshold, calculating all frequent item sets, and returning a final frequent item set and the confidence coefficient through iteration to form an initial association rule.
In some embodiments, the step II comprises:
1) optimizing the product combination obtained by the FP-tree algorithm through the historical product combination probability to form a final association rule;
2) selecting a test enterprise, standardizing the purchased products of the enterprise, matching the products by an association rule engine, and outputting the associated products.
In some embodiments, the step III comprises:
1) based on the product portrait and the enterprise portrait, calculating the similarity of the product portrait and the enterprise portrait through cosine similarity
Figure BDA0002513073360000051
In the formula (6), xiFor enterprise portrait vectors (attribute values measured by 0, 1), yiIs the product portrait vector (0, 1 is used to measure the attribute value), cos (theta) is the similarity between the product portrait and the enterprise portrait;
2) and sorting the similarity calculation results, and recommending the products according to a set threshold value.
In some embodiments, the step IV comprises:
a) storing the standardized enterprise names, the recommended products obtained by calculation in the step III and the corresponding probabilities into a database to form an original business opportunity recommendation table;
b) and establishing a filtering catalog to form a Filter, processing the original business opportunity recommendation table by using the Filter to form a final recommendation business opportunity library, wherein the filtering catalog comprises an industry Filter catalog and a recommendation language Filter catalog.
Compared with the prior art, the invention provides a digital and omnibearing platform operation service system construction method integrating the platform service input end and the platform service output end into a whole by taking the algorithm logic of ' total enterprise amount introduction rate ' life cycle business value ' as value guidance and taking big data and intelligent technology as the basis, thereby realizing benign operation and deepened application of the platform.
Drawings
Fig. 1 is a schematic diagram of a method for constructing a platform operation service system according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an image processing label system according to an embodiment of the present invention.
Detailed Description
As described above, in view of the deficiencies of the prior art, the present inventors have made extensive research and practice to provide a technical solution of the present invention, and specifically, provide a method for constructing a digital and comprehensive platform operation service system based on an improved algorithm logic of "total platform operation value is total enterprise amount, introduction rate, life cycle business value", wherein core engines such as enterprise portrait, business mining, etc. are applied based on big data and intelligent technology to construct system modules such as service portrait, credit management, business mining, service recommendation, etc. by technically enabling a service provider, the full life cycle value of the service enterprise is deeply mined by continuous service, a bidirectional feedback mechanism is established, and iterative upgrade is performed continuously, thereby realizing benign operation and deep application of the platform.
The present invention will be described in further detail with reference to the following examples and drawings, which are intended to facilitate the understanding of the present invention and are not intended to limit the present invention in any way. Unless otherwise specified, ETL tools, machine learning algorithms, etc. employed in the following embodiments are known in the art.
Referring to fig. 1, a method for constructing a platform operation service system provided in this embodiment includes the following steps:
the method comprises the following steps: and establishing a platform big data base comprising industrial and commercial data, license data, intellectual property data, public opinion data, policy data and the like of enterprise users, service providers and the like. The method comprises the steps of firstly, acquiring data, obtaining various types of structured, semi-structured and unstructured mass data based on social network data, mobile internet data and other modes, then putting the data into Hive through ETL tools such as Kettle, Sqoop, flash and the like, processing and processing the data by using Mapreduce and Spark, and constructing a platform big data base.
Step two: and constructing a service provider management system which comprises functional modules of service provider portrait, service provider credit evaluation and the like.
The integrated comprehensive service platform has a lot of service provider data, how to automatically connect high-quality service providers for clients in the platform operation process, reduce intermediate cost and form virtuous circle, and depends on the platform to accurately grasp the service provider information to a great extent. The service provider management system set up in the second step is additionally provided with functional modules such as service provider portrait, service provider retrieval, service provider credit evaluation and the like besides a service provider access auditing system module, and also provides a service provider client side, so that business opportunity push and effectiveness feedback are realized, and instant communication and feedback of service providers and platforms are facilitated.
Furthermore, the platform operation of this embodiment needs to manage the service providers, and with the service provider admission audit system module as a starting point, a service provider image module basis is established, the service provider audit module and the service provider credit evaluation module are applied, so that important links such as platform user service provider selection and service intelligent matching are realized, and in addition, the service provider client provides functions such as audit, information submission, feedback and the like for a plurality of service providers.
Furthermore, the service provider admission auditing module is a starting point for platform operation service provider management, and is used for setting service provider admission standards and index thresholds, including service provider establishment years, registered capital, enterprise scale, whether more service disputes exist, whether to enter a loss of credit blacklist and the like, so as to realize warehousing and management of service providers;
the service provider portrait module is used for portrait processing from the aspects of basic information, delivery capacity, service level and the like, and is the basis of service provider retrieval, credit evaluation and service matching;
the service provider retrieval module is used for collecting service providers according to characteristics such as large enterprise scale and no bad record according to the processing result of the service provider portrait module;
the service provider credit evaluation module is used for performing credit evaluation based on AHP and fuzzy comprehensive evaluation method according to the processing result of the service provider image module and a credit evaluation index system (innovation development capability status, service level, credit history, performance capability, financial status and management analysis), and the process comprises the following steps:
A. constructing a relevant judgment matrix of a credit evaluation index system, and constructing a judgment matrix X ═ Xij](i, j ═ 1, 2, 3.., n), where X isij＝Xi/XjMeans XiRelative to XjThe importance of the judgment matrix is constructed in turn for each level,
Figure BDA0002513073360000071
B. determining the credit evaluation of a target layer service provider, the relative weight coefficients of a secondary index layer (innovation development capability condition, service level, credit history, performance capability, financial condition, management analysis and the like) and a specific measurement index layer (such as the number of people invested in research and development, the number of invested funds in research and development, the number of mobile funds, the number of people managing teams and the like) by the following steps:
XW＝λmaxw type (1)
Figure BDA0002513073360000072
Figure BDA0002513073360000073
W＝(w1，w2，…，wn)TFormula (4)
In the formula (1), X is a judgment matrix, W is a weight vector, and lambdamaxIn order to determine the maximum eigenvalue of the matrix,
x in the formula (2)ijIs an index xiRelative to xjThe importance of (a) to (b),
Figure BDA0002513073360000074
to measure specifically the weight of the index layer relative to the secondary index layer,
in the formula (3)
Figure BDA0002513073360000075
Also specifically measures the weight, w, of the index layer relative to the secondary index layeriIs the weight of the secondary index relative to the target layer,
in the formula (4), W is also a weight vector;
C. carrying out consistency check on the calculation model, calculating a model consistency index CI, determining a corresponding random consistency index RI, calculating a consistency ratio CR, and continuously correcting the model until CR is less than 0.1, so as to meet the requirements, wherein:
Figure BDA0002513073360000081
in formula (5) < lambda >maxIn order to judge the maximum eigenvalue of the matrix, n is the number of calculation indexes;
step three: and constructing an introduced customer flow pool, forming a transaction database, and cleaning and processing data to construct a product portrait and an enterprise portrait.
Since it is the starting point to recognize the characteristics of the enterprise users, it is important to construct an enterprise portrait engine, and in the prior art, portraits are mostly constructed from both static information data and dynamic information data.
The embodiment builds an enterprise sketch engine, and adopts a systematic analysis framework, namely: the enterprise is helped to introduce a large amount of customer traffic based on multi-channel API docking, promotion activities and the like, and an enterprise and a transaction data warehouse (namely, the transaction database) are formed through introducing strategy rule judgment. The introduction strategy is mainly used for screening enterprise users, and comprises channel classification marking, business abnormality and risk enterprise elimination, service history blacklist enterprise elimination and the like, in the process, enterprise ID is established, a transaction database is enriched continuously, and the process specifically comprises the following steps:
the channel classification mark is used for establishing source and preference basic data of enterprise users, a specified channel enterprise user group can be directly selected at a specific time point, and a flow pool is formed by multi-channel API docking and promotion activities, so that the enterprise user data has various types and inconsistent information such as cross repetition, and the channel classification mark can help to introduce and manage a transaction database;
and secondly, eliminating the abnormal operation and the enterprise with the risk, and improving the safety and the effectiveness of the transaction database by filtering enterprise users with higher risk or abnormal operation.
And thirdly, service history blacklist enterprises are removed, potential problems that enterprise users do not perform appointments, arrears and the like cannot be avoided in the platform operation process, so that the service history blacklist enterprises need to be further removed in the introduction process, and the platform operation is ensured to be carried out orderly.
The enterprise ID is the unique identification code of the enterprise user of the platform, and the problem of name non-standardization of platform enterprise user storage, name abbreviation and the like is possibly caused due to different channel sources and the like, so that subsequent data analysis is influenced, and therefore, the unique identification code (enterprise ID) of the enterprise user is established by data cleaning, namely comparing, matching, text processing and the like of the enterprise name in the business data of different channel sources in the flow pool with the business name library, and the data validity and the accuracy of the algorithm and the model are improved.
On the basis of the transaction database, a multi-dimensional image is constructed through a rule engine and a machine learning algorithm from the aspects of a basic label, an operation label, a product label, a risk label and the like, and an enterprise portrait model and a product portrait model are continuously optimized by combining an expert feedback and a weight model. The specific process is as follows:
s1: establishing an image processing label system including aspects such as a basic label, an operation label, a product label and a risk label, for example, refer to the following fig. 2;
s2: forming an enterprise portrait and product portrait feature lexicon based on data foundations of industrial and commercial data, intellectual property data and the like of enterprise users, wherein the established age portrait label in the basic label comprises a new company, a young company, a very young company, a mature company and a long-life company;
s3: selecting an enterprise sample, and matching and processing an enterprise portrait based on a regular expression according to a portrait processing rule base (Image _ Process _ Rules);
s4: aiming at a transaction standardized product, establishing a product portrait processing rule which comprises information of product suitability for industry, region, scale, age and the like;
s5: and selecting enterprise IDs in the same industry and the portrait processing result in the step S3, obtaining similar enterprises based on a K-Means clustering algorithm, and further optimizing the portrait processing rule.
Step four: and constructing a business mining basic engine and a business base, wherein the business mining basic engine comprises a multi-source fusion repeated purchasing business engine based on an association rule algorithm and a similarity algorithm and the like.
The business opportunity mining is a key point for associating service providers and enterprise users and is a basic point for platform operation, the business opportunity value of the life cycle is based on the algorithm logic of the platform operation value, the repurchase is continuously generated, and a product association recommendation algorithm model based on association rules is constructed to construct a repurchase business opportunity engine. The fourth step may specifically include the following steps:
i: FP-tree algorithm based associated product rule and probability thereof obtained from historical transaction data
1) Selecting enterprise samples, and acquiring all historical transaction records based on enterprise IDs
2) Based on a standardized service product name library, applying standardized product names to standardize purchased products of sample enterprises, forming an original product combination according to transaction time, and generating a training data set;
3) setting a support threshold, calculating all frequent item sets, and returning a final frequent item set and a confidence coefficient through iteration to form an initial association rule;
II: finding out the associated products of the purchased products according to the association rules;
optimizing the product combination obtained by the FP-tree algorithm through the historical product combination probability to form a final association rule, such as (trademark registration → 400 telephone);
selecting a test enterprise, standardizing the purchased products of the enterprise, matching the products by an association rule engine, and outputting the associated products;
III: calculating the similarity of the product portrait and the enterprise portrait, sorting according to the similarity calculation result, and finding a recommended product according to a set threshold;
i) based on the product portrait and the enterprise portrait processed in the third step, the similarity between the product portrait and the enterprise portrait is calculated through cosine similarity (formula below)
Figure BDA0002513073360000101
X in formula (6)iFor enterprise portrait vectors (attribute values measured by 0, 1), yiIs the product portrait vector (0, 1 is used to measure the attribute value), cos (theta) is the similarity between the product portrait and the enterprise portrait;
and ii) sorting the similarity calculation results, and recommending products according to a set threshold value.
IV: and filtering the products by using a filter, and combining the enterprises and the products to form a recommender hangar.
a) Storing the standardized enterprise name, the recommended product calculated according to the image similarity and the probability thereof into a database to form an original business opportunity recommendation table;
b) establishing catalogues such as an industry Filter and a recommended technology Filter to form a Filter Filter, and processing the original business opportunity recommendation table obtained in the step a) to form a final business opportunity recommendation library.
Step five: building a service recommendation system, and carrying out intelligent service matching and business process supervision
In the step, potential business opportunities of enterprise users are deeply mined on the basis of a business opportunity mining basic engine and a transaction database, the potential business opportunities are automatically pushed to matched service providers through an intelligent matching algorithm, early warning and supervision are carried out through a full-process monitoring technology, and the effectiveness and deep application of platform operation are improved to the greatest extent.
Step six: constructing a bidirectional feedback mechanism
In the step, intelligent recommendation algorithm and system optimization can be realized according to the feedback of the service provider on business opportunity effectiveness, and the service provider portrait and the management system thereof can be optimized according to the evaluation feedback of the service enterprise. Through a bidirectional feedback mechanism continuous optimization algorithm, accurate business opportunity mining and butt joint are realized, the operation labor cost is further released, the experience decision of operation is replaced by the data logic of a machine, an intelligent operation system is really formed, and therefore the maximum platform operation value is achieved.
The construction method of the platform operation service system provided by the embodiment of the invention takes the logical algorithm of 'total enterprise amount, introduction rate and life cycle business value' as the value guide, changes the operation thinking as a whole, has subversive innovation, and simultaneously builds the omnibearing integrated service facing the life cycle of the service enterprise based on digital drive, has high integration and expansibility, can increase the client viscosity, promote and deeply apply the power-assisted platform, and fully exerts the value of data assets, wherein the adoption of the intelligent service matching and the bidirectional feedback mechanism can also continuously optimize the accuracy of data processing, and ensure the high efficiency of the system and the flexibility of the system.
In the algorithm logic for constructing the support by adopting the platform operation service system constructed by the embodiment of the invention, the introduction rate is improved at low cost by taking digitalization as support and passing through the total amount of enterprises in the flow pool, then business mining is carried out based on similarity calculation, association algorithm and the like, continuous repeated purchase of service enterprises is realized, and the service input end level is continuously enabled and optimized by a service provider management system such as feedback, service monitoring and the like, so that the continuous increase of service quality, service transaction value and the like is formed.
It should be understood that the technical solution of the present invention is not limited to the above-mentioned specific embodiments, and all technical modifications made according to the technical solution of the present invention fall within the protection scope of the present invention without departing from the spirit of the present invention and the protection scope of the claims.

Claims (10)

1. A method for constructing a platform operation service system is characterized by comprising the following steps:
collecting data to construct a basic database, wherein the collected data comprises any one or combination of more of industrial and commercial data, license data, intellectual property data, public opinion data and policy data;
constructing a service provider management system, which comprises a service provider admission auditing module, a service provider imaging module, a service provider retrieving module, a service provider credit evaluation module and a service provider client;
establishing an introduced customer flow pool, forming a transaction database, and then establishing a product portrait and an enterprise portrait of an enterprise user;
providing a business mining basic engine, wherein the business mining basic engine is a multi-source fusion repeated purchasing business engine based on an association rule algorithm and a similarity algorithm;
constructing a business mining system by utilizing the business mining basic engine, the basic database and the transaction database, and intelligently matching the business mining system with the service provider management system;
potential business opportunities of enterprise users are mined by the business opportunity mining system, the potential business opportunities are automatically pushed to the matched service providers, and the business opportunity mining system is optimized according to business opportunity effectiveness evaluation fed back by the matched service providers;
and enabling the matched service provider to provide services for the corresponding selected enterprise users according to the potential business opportunities, and enabling the service provider management system to optimize according to the service evaluation fed back by the selected enterprise users.
2. The building method according to claim 1, characterized by comprising: obtaining various types of structured, semi-structured and unstructured mass data at least based on social network data and mobile internet data, then placing the mass data into Hive through an ETL tool, processing and processing the data by using Mapreduce and Spark, and constructing a platform big data base, namely obtaining the basic database.
3. The building method according to claim 1, characterized by comprising:
the service provider admission auditing module is used for realizing warehousing and management of service providers by setting service provider admission standards and index thresholds, wherein the service provider admission standards and index thresholds comprise service provider establishment years, registered capital, enterprise scale, whether more service disputes exist or not and whether the service providers enter a loss-of-trust blacklist or not;
the service provider portrait module is at least used for portrait processing from basic information, delivery capacity and service level of a service provider;
the service provider retrieval module is at least used for collecting service providers according to the processing result of the service provider image module and the service provider characteristics, wherein the service provider characteristics comprise that an enterprise achieves a set scale and has no bad records;
the service provider credit evaluation module is used for performing credit evaluation according to the processing result of the service provider image module and a credit evaluation index system based on AHP and a fuzzy comprehensive evaluation method, wherein the credit evaluation index system comprises innovation development capability status, service level, credit history, performance capability, financial status and management analysis;
the service provider client is at least used for providing a port for auditing, information submitting and feedback for a service provider;
the process of credit evaluation based on the AHP and the fuzzy comprehensive evaluation method comprises the following steps:
A. constructing a relevant judgment matrix of a credit evaluation index system, and constructing a judgment matrix X ═ Xij](i, j ═ 1, 2, 3.., n), where X isij＝Xi/XjMeans XiRelative to XjThe importance of the judgment matrix is sequentially constructed for each level;
Figure FDA0002513073350000021
B. according to the following formulas 1) to 4), determining relative weight coefficients of a target layer service provider credit evaluation, a secondary index layer and a specific measurement index layer, wherein the secondary index layer comprises innovation development capability status, service level, credit history, performance capability, financial status and management analysis, and the specific measurement index layer comprises research and development invested number, liquidity and management team number;
XW＝λmaxw type (1)
Figure FDA0002513073350000022
Figure FDA0002513073350000023
W＝(w1，w2，…，wn)TFormula (4)
X in the formula (1) is a judgment matrix and lambdamaxTo determine the maximum eigenvalue of the matrix, x in equation (2)ijIs an index xiRelative to xjOf importance, w in formula (3)iW is a weight vector in the formula (1) and the formula (4), and W is a weight vector in the formula (2) and the formula (3)
Figure FDA0002513073350000024
The weight of the index layer relative to the second-level index layer is measured specifically;
C. carrying out consistency check on the calculation model according to the following formula 5), calculating a model consistency index CI, determining a corresponding random consistency index RI, calculating a consistency ratio CR, and continuously correcting the model until the CR is less than 0.1, wherein:
Figure FDA0002513073350000025
wherein λ ismaxTo determine the maximum eigenvalue of the matrix, n is the number of calculation indices.
4. The building method according to claim 1, characterized by comprising: introducing a large amount of customer flow into enterprise users at least through multi-channel API docking and promotion activities to form an introduced customer flow pool, judging through introduced strategy rules to form a transaction database, and then cleaning and processing data to construct product figures and enterprise figures of the enterprise users;
the introduction strategy is at least used for screening enterprise users through channel classification marking, abnormal operation and risk enterprise elimination and service history blacklist enterprise to establish enterprise ID and enrich a transaction database;
the channel classification mark is used for establishing source and preference basic data of enterprise users so as to help introduce and manage a transaction database;
the enterprise with the removed operation abnormity and the risk is used for improving the safety and the effectiveness of the transaction database by filtering enterprise users with higher risk or operation abnormity;
the enterprise with the service history rejection blacklist is used for rejecting the enterprise with bad service history;
the enterprise ID is a unique identification code of an enterprise user, and is at least established by comparing, matching and text processing enterprise names in business data of different channel sources in the flow pool with a business name library.
5. The building method according to claim 4, characterized by comprising: based on the transaction database and the portrait processing label system, a multi-dimensional portrait is constructed through a rule engine and a machine learning algorithm, an enterprise portrait and a product portrait model of an enterprise user are optimized by combining expert feedback and a weight model, and then the product portrait and the enterprise portrait of all the enterprise users are processed by utilizing the optimized enterprise portrait and the optimized product portrait model.
6. The building method according to claim 5, characterized by comprising:
s1: establishing an image processing label system, wherein the image processing label system comprises a basic label, an operation label, a product label and a risk label;
s2: forming an enterprise portrait and a product portrait feature word bank based on basic data, wherein the basic data comprises industrial and commercial data and intellectual property data of enterprise users;
s3: selecting an enterprise sample, and matching and processing the enterprise portrait based on a regular expression according to the portrait processing rule base;
s4: aiming at a transaction standardized product, establishing a product portrait processing rule, wherein the product portrait processing rule comprises that the product is suitable for the industry, suitable regions, scale and age;
s5: and selecting enterprise IDs in the same industry and the portrait processing result in the step S3, obtaining similar enterprises based on a K-Means clustering algorithm, and further optimizing the portrait processing rule.
7. The building method according to claim 1, characterized by comprising:
I. acquiring a related product rule and a corresponding probability from historical transaction data based on an FP-tree algorithm;
II. Finding out the associated products of the purchased products according to the association rules;
III, calculating the similarity of the product portrait and the enterprise portrait, sorting according to the similarity calculation result, and finding a recommended product according to a set threshold;
and IV, filtering the products by using a filter, and combining the enterprises and the products to form a recommender library.
8. The construction method according to claim 7, wherein step I comprises:
1) selecting an enterprise sample, and acquiring all historical transaction records based on enterprise IDs;
2) based on the standardized service product name list, standardizing the purchased products of the sample enterprise, forming an original product combination according to the transaction time, and generating a training data set;
3) setting a support threshold, calculating all frequent item sets, and returning a final frequent item set and the confidence coefficient through iteration to form an initial association rule.
9. The construction method according to claim 8, wherein step II comprises:
1) optimizing the product combination obtained by the FP-tree algorithm through the historical product combination probability to form a final association rule;
2) and selecting a test enterprise, standardizing the purchased products of the enterprise, matching the final association rules, and outputting the associated products.
10. The construction method according to claim 9,
the step III comprises the following steps:
1) based on the product portrait and the enterprise portrait, calculating the similarity of the product portrait and the enterprise portrait through cosine similarity
Figure FDA0002513073350000041
Wherein x isiFor enterprise portrait vectors, yiIs a product image vector, cos (θ) is the similarity between the product image and the enterprise image, xi、yiThe attribute values are both measured by 0 and 1;
2) sorting the similarity calculation results, and recommending products according to a set threshold;
the step IV comprises the following steps:
a) storing the standardized enterprise names, the recommended products obtained by calculation in the step III and the corresponding probabilities into a database to form an original business opportunity recommendation table;
b) and establishing a filtering catalog to form a Filter, processing the original business opportunity recommendation table by using the Filter to form a final recommendation business opportunity library, wherein the filtering catalog comprises an industry Filter catalog and a recommendation language Filter catalog.
CN202010471261.7A 2020-05-28 2020-05-28 Construction method of platform operation service system Active CN111612549B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010471261.7A CN111612549B (en) 2020-05-28 2020-05-28 Construction method of platform operation service system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010471261.7A CN111612549B (en) 2020-05-28 2020-05-28 Construction method of platform operation service system

Publications (2)

Publication Number Publication Date
CN111612549A true CN111612549A (en) 2020-09-01
CN111612549B CN111612549B (en) 2023-07-21

Family

ID=72198615

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010471261.7A Active CN111612549B (en) 2020-05-28 2020-05-28 Construction method of platform operation service system

Country Status (1)

Country Link
CN (1) CN111612549B (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112100237A (en) * 2020-09-04 2020-12-18 北京百度网讯科技有限公司 User data processing method, device, equipment and storage medium
CN112232909A (en) * 2020-10-13 2021-01-15 汉唐信通(北京)科技有限公司 Business opportunity mining method based on enterprise portrait
CN112330047A (en) * 2020-11-18 2021-02-05 交通银行股份有限公司 Credit card repayment probability prediction method based on user behavior characteristics
CN112488639A (en) * 2020-11-12 2021-03-12 深圳市中博科创信息技术有限公司 Construction method of enterprise service system based on full life cycle
CN112508425A (en) * 2020-12-14 2021-03-16 东南大学 Method for constructing city trip user portrait system for flexible public transportation system
CN113377742A (en) * 2021-06-02 2021-09-10 浪潮软件股份有限公司 Corporate spatial data application method based on corporate comprehensive data resource library
CN114491265A (en) * 2022-01-28 2022-05-13 北京乐开科技有限责任公司 Construction method of operation service system of business space platform
CN114757797A (en) * 2022-06-13 2022-07-15 国网浙江省电力有限公司 Power grid resource service central platform architecture method based on data model drive
TWI795809B (en) * 2021-06-17 2023-03-11 華南商業銀行股份有限公司 Business evaluation system and method therefore
CN117149799A (en) * 2023-11-01 2023-12-01 建信金融科技有限责任公司 Data updating method, device, electronic equipment and computer readable medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108197997A (en) * 2018-01-16 2018-06-22 成都小时代科技有限公司 A kind of automobile shops inside trade machine excavates and the method and system of conversion
CN108241932A (en) * 2018-01-24 2018-07-03 国网山东省电力公司泰安供电公司 A kind of method for building up of electricity provider evaluation model
CN109767255A (en) * 2018-12-06 2019-05-17 东莞团贷网互联网科技服务有限公司 A method of it is modeled by big data and realizes intelligence operation and precision marketing
CN111080440A (en) * 2019-12-18 2020-04-28 上海良鑫网络科技有限公司 Big data wind control management system

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108197997A (en) * 2018-01-16 2018-06-22 成都小时代科技有限公司 A kind of automobile shops inside trade machine excavates and the method and system of conversion
CN108241932A (en) * 2018-01-24 2018-07-03 国网山东省电力公司泰安供电公司 A kind of method for building up of electricity provider evaluation model
CN109767255A (en) * 2018-12-06 2019-05-17 东莞团贷网互联网科技服务有限公司 A method of it is modeled by big data and realizes intelligence operation and precision marketing
CN111080440A (en) * 2019-12-18 2020-04-28 上海良鑫网络科技有限公司 Big data wind control management system

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112100237A (en) * 2020-09-04 2020-12-18 北京百度网讯科技有限公司 User data processing method, device, equipment and storage medium
CN112100237B (en) * 2020-09-04 2023-08-15 北京百度网讯科技有限公司 User data processing method, device, equipment and storage medium
CN112232909A (en) * 2020-10-13 2021-01-15 汉唐信通(北京)科技有限公司 Business opportunity mining method based on enterprise portrait
CN112488639A (en) * 2020-11-12 2021-03-12 深圳市中博科创信息技术有限公司 Construction method of enterprise service system based on full life cycle
CN112330047A (en) * 2020-11-18 2021-02-05 交通银行股份有限公司 Credit card repayment probability prediction method based on user behavior characteristics
CN112508425A (en) * 2020-12-14 2021-03-16 东南大学 Method for constructing city trip user portrait system for flexible public transportation system
CN112508425B (en) * 2020-12-14 2024-03-15 东南大学 Urban travel user portrait system construction method for elastic public transport system
CN113377742A (en) * 2021-06-02 2021-09-10 浪潮软件股份有限公司 Corporate spatial data application method based on corporate comprehensive data resource library
TWI795809B (en) * 2021-06-17 2023-03-11 華南商業銀行股份有限公司 Business evaluation system and method therefore
CN114491265B (en) * 2022-01-28 2022-08-23 北京乐开科技有限责任公司 Construction method of operation service system of business space platform
CN114491265A (en) * 2022-01-28 2022-05-13 北京乐开科技有限责任公司 Construction method of operation service system of business space platform
CN114757797A (en) * 2022-06-13 2022-07-15 国网浙江省电力有限公司 Power grid resource service central platform architecture method based on data model drive
CN117149799A (en) * 2023-11-01 2023-12-01 建信金融科技有限责任公司 Data updating method, device, electronic equipment and computer readable medium
CN117149799B (en) * 2023-11-01 2024-02-13 建信金融科技有限责任公司 Data updating method, device, electronic equipment and computer readable medium

Also Published As

Publication number Publication date
CN111612549B (en) 2023-07-21

Similar Documents

Publication Publication Date Title
CN111612549B (en) Construction method of platform operation service system
CN112232909A (en) Business opportunity mining method based on enterprise portrait
CN110704572A (en) Suspected illegal fundraising risk early warning method, device, equipment and storage medium
CN104424613A (en) Value added tax invoice monitoring method and system thereof
CN113779264A (en) Trade recommendation method based on patent supply and demand knowledge graph
CN114860916A (en) Knowledge retrieval method and device
CN116645129A (en) Manufacturing resource recommendation method based on knowledge graph
Cetindamar et al. Assessing big data analytics capability and sustainability in supply chains
CN117217634B (en) Enterprise cooperation community discovery method based on complex network
CN112784049B (en) Text data-oriented online social platform multi-element knowledge acquisition method
Fucheng Financial performance of intelligent manufacturing enterprises based on fuzzy neural network and data twinning
Li Data quality and data cleaning in database applications
WO2020253353A1 (en) Resource acquisition qualification generation method for preset user and related device
CN115345401A (en) Six-dimensional analysis method for finding enterprise financial risk
CN117112794A (en) Knowledge enhancement-based multi-granularity government service item recommendation method
Long [Retracted] Analysis of Insurance Marketing Planning Based on BD‐Guided Decision Tree Classification Algorithm
Feng et al. Evaluation of dynamic technological innovation capability in high-tech enterprises based on pythagorean fuzzy LBWA and MULTIMOORA
CN113379211A (en) Block chain-based logistics information platform default risk management and control system and method
Azeroual et al. Overlooked Aspects of Data Governance: Workflow Framework For Enterprise Data Deduplication
Assaf et al. RUBIX: a framework for improving data integration with linked data
Yang Automatic Decision Algorithm of Interpretation Power in Criminal Justice Based on Data Activity Consultant
Sun Management Research of Big Data Technology in Financial Decision-Making of Enterprise Cloud Accounting
Bai et al. Research on Audit Data Analysis and Decision Tree Algorithm for Benefit Distribution of Enterprise Financing Alliance
Wang Analysis and evaluation of engineering job demand based on big data technology
Bai The application of customer relationship management and data mining in Chinese insurance companies

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