CN114066572A - Cable transaction intelligent recommendation method and system - Google Patents

Cable transaction intelligent recommendation method and system Download PDF

Info

Publication number
CN114066572A
CN114066572A CN202111364667.6A CN202111364667A CN114066572A CN 114066572 A CN114066572 A CN 114066572A CN 202111364667 A CN202111364667 A CN 202111364667A CN 114066572 A CN114066572 A CN 114066572A
Authority
CN
China
Prior art keywords
cable
client
data
recommendation
cables
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
CN202111364667.6A
Other languages
Chinese (zh)
Other versions
CN114066572B (en
Inventor
栾小丽
许飞鸿
杜康
贾明
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.)
Jiangsu Wuyuntong Logistics Technology Co ltd
Jiangnan University
Original Assignee
Jiangsu Wuyuntong Logistics Technology Co ltd
Jiangnan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangsu Wuyuntong Logistics Technology Co ltd, Jiangnan University filed Critical Jiangsu Wuyuntong Logistics Technology Co ltd
Priority to CN202111364667.6A priority Critical patent/CN114066572B/en
Publication of CN114066572A publication Critical patent/CN114066572A/en
Application granted granted Critical
Publication of CN114066572B publication Critical patent/CN114066572B/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
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9536—Search customisation based on social or collaborative filtering

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses an intelligent recommendation method and system for cable transaction, which are characterized by quantizing cable model parameters and requirements of customers on different scene applications based on a recommendation algorithm of content and collaborative filtering, and converting manual research and big data screening into intelligent recommendation, thereby solving the problem that the customers select cables with time and labor consumption, and simultaneously, because the system considers the differentiation of new and old customers and solves the difficult problem of cold start, the system can stably and quickly recommend proper cables for all the customers, thereby promoting the sale of cable manufacturers, improving the convenience of purchasing cables by the customers, secondly, for the situation that the recommendation satisfaction degree is low due to the fact that the historical data quantity of certain cables is insufficient in the recommendation process, the intelligent expansion network is used for expanding the data and training the model, so that a better recommendation effect is obtained, the recommendation accuracy is improved, and the recommendation problem of small-sample cable data is solved.

Description

Cable transaction intelligent recommendation method and system
Technical Field
The invention relates to the technical field of big data, in particular to a cable transaction intelligent recommendation method and system.
Background
Cables are generally used for transmitting electric energy and signals, are widely used in the fields of electric power systems, information transmission systems, mechanical devices, instrument and meter systems, and the like, and are indispensable goods. When a customer selects a cable, not only the model parameters and quality of the cable need to be considered, but also the scale, the region to which the cable belongs, the service attitude and the like of a manufacturer need to be weighed, the selection and evaluation of the cable are not complex tasks, but the cable trading market is huge, the variety and the quality of the cable are different, and how to quickly and efficiently select a proper cable by the customer is a challenging problem.
The current mode of enterprise cable transaction selection mainly has 2: firstly, manual investigation is carried out, and a client contacts different cable manufacturers to investigate the condition of cables sold by the manufacturers and the qualification of the manufacturers. The method consumes manpower and material resources, has low efficiency, and the investigation result is often influenced by the subjectivity of the investigation personnel and is gradually eliminated at present; secondly, the cable is selected through information screening under big data, although the method is rapid, the screening standard is only the requirement of a customer on the cable, the cable meeting the requirement of the customer is difficult to obtain accurately, the customer still needs to spend a great deal of energy to further investigate the commodity and a manufacturer thereof, and the efficiency of the whole transaction process is low.
In recent years, with the continuous development of data mining and machine learning, a plurality of recommendation systems for helping users to quickly select commodities emerge, therefore, the invention provides an intelligent recommendation method and system for cable transactions, which utilize a recommendation algorithm based on content and collaborative filtering emerging in recent years, use a special label of the cable commodity and a special user portrait in quick recommendation of cables, screen the quick intelligent recommendation from common big data information, and make up for the shortage that manpower and material resources are consumed in the transaction process. The invention considers the difference of the new and old customers and solves the problem of cold start, and the system can stably and quickly recommend proper cables for all the customers, thereby promoting the sale of cable manufacturers and improving the convenience of purchasing cables by the customers.
Disclosure of Invention
The invention aims to provide a cable transaction intelligent recommendation method which is more efficient, more accurate and high in recommendation efficiency.
In order to solve the above problems, the present invention provides an intelligent cable transaction recommendation method, which includes the following steps:
s1, acquiring and sorting information of the cleaned cable, and establishing a commodity database;
s2, acquiring basic information of the client, extracting historical behavior data of the client and establishing a client database;
s3, judging whether the current client is a new client, if so, executing a step S4, otherwise, executing a step S8;
s4, calling the basic information of the new client in the client database, and obtaining the neighbor client through similarity comparison;
s5, deducing one or more application scenes of the new client application cable according to the transaction requirements of the neighbor clients;
s6, calling cable parameters and manufacturer information in the commodity database according to the application scene to perform adaptability calculation to obtain adaptability scores corresponding to different cables;
s7, recommending one or more cables with the scores higher than those of the cables according to different application scenes, recording behavior feedback of a new client on a recommendation result, and updating a client database in real time;
s8, calling behavior data of the client in the client data, and calculating the degree of intention of the client to different cables through a collaborative filtering algorithm;
and S9, screening cables which may be interested by the customer according to the popularity, recording behavior data of the customer in the transaction process, and updating the customer database in real time.
As a further improvement of the present invention, between step S8 and step S9, the following steps are further included:
A. and collecting the cable data with the least samples of the same type in the data, modeling and training by using an intelligent expansion network, obtaining a recommendation result with the highest similarity, and calculating the degree of intention of a client on the sample cable.
As a further improvement of the invention, step A comprises:
a1, establishing two weight-sharing sub-networks with the same structure, inputting two groups of cable data X1And X2Respectively convert it into vector Gw(X1) And Gw(X2) Then, the distance E of the two vectors is calculated by a distance measurement methodw；
A2, assuming that the input sample of the intelligent expansion network is (X)1,X2Y), construct the loss function as follows:
Figure BDA0003360174800000031
L(W,(X1,X2,y)i)＝(1-y)LG(Ew(X1,X2)i)+yLI(Ew(X1,X2)i)
wherein (X)1,X2,y)iIs the ith input sample, where LGTo calculate the loss function for the cable for the same class only, LICalculating loss functions of different categories to the cable; the label y-0 indicates two sets of cable data X1And X2Of different types, the label y ═ 1 denotes two sets of cable data X1And X2Belonging to the same type, the setting of its loss function follows: when y is 0, the greater the distance, the smaller the loss, i.e. with respect to EwWhen y is 1, the smaller the distance, the greater the loss, i.e. with respect to EwIs a monotonically increasing function of.
A3, mixing LGSet to monotonically increase, set L to monotonically increaseIAnd setting the data to be monotonously reduced, and training a loss function by using the cable data with the least samples of the same type in the data.
As a further improvement of the invention, in the obtaining of the neighbor clients by similarity comparison, the similarity w between two clientsuvComprises the following steps:
Figure BDA0003360174800000032
wherein the scoring vectors of the two clients u and v are respectively
Figure BDA0003360174800000033
And
Figure BDA0003360174800000034
as a further improvement of the present invention, the suitability scores corresponding to the different cables are as follows:
Figure BDA0003360174800000035
where C is the nearest neighbor set of the new customer u, Rv,jScoring Cable j for neighboring customer v, Pu,jThe new customer is scored for the prediction of cable j.
As a further improvement of the invention, the cable information comprises one or more of materials, core number, current-carrying capacity, insulation level, authentication, application scenario, manufacturer scale, manufacturer location and manufacturer service level, and further comprises insulation materials and sheath, armor and bus type.
As a further improvement of the present invention, the customer basic information includes one or more of name, gender, age, affiliated entity, position and shipping address.
As a further improvement of the invention, the historical behavior data comprises one or more of search keywords, browsing records, purchasing records, communication records, collecting records and evaluation records.
As a further improvement of the invention, the cleaning comprises one or more of text error correction, keyword extraction, meaning analysis, feature coding and data standardization.
In order to solve the above problem, the present invention further provides an intelligent cable transaction recommendation system, which includes:
the commodity database establishing module is used for acquiring and arranging the cable cleaning information and establishing a commodity database;
the client database establishing module is used for acquiring the basic information of the client, extracting the historical behavior data of the client and establishing a client database;
the judging module is used for judging whether the current client is a new client or not;
the similarity comparison module is used for calling the basic information of the new client in the client database and acquiring the neighbor client through similarity comparison;
the application scenario inference module is used for inferring one or more application scenarios of a new client application cable according to the transaction requirements of the neighbor clients;
the adaptability calculation module is used for calling cable parameters and manufacturer information in the commodity database according to the application scene to perform adaptability calculation to obtain adaptability scores corresponding to different cables;
the cable recommendation module is used for recommending one or more cables with the grades earlier according to different application scenes, recording the behavior feedback of a new client on a recommendation result, and updating a client database in real time;
the system comprises a popularity calculation module, a collaborative filtering module and a data processing module, wherein the popularity calculation module is used for calling behavior data of a client in client data and calculating the popularity of the client to different cables through a collaborative filtering algorithm;
and the cable screening module is used for screening cables possibly interested by the customer according to the popularity, recording behavior data of the customer in the transaction process and updating the customer database in real time.
The invention has the beneficial effects that:
the cable transaction intelligent recommendation method and system quantize the cable model parameters and the requirements of the customers on different scene applications into characteristics based on the recommendation algorithm of content and collaborative filtering, and convert manual research and big data screening into intelligent recommendation, so that the problem that the customers select cables with time and labor consumption is solved.
Meanwhile, the system considers the differentiation of new and old customers and solves the problem of cold start, and the system can stably and quickly recommend proper cables to all the customers, so that the sale of cable manufacturers is promoted, and the convenience of purchasing the cables by the customers is improved.
Secondly, aiming at the condition that the recommendation satisfaction degree is low due to the fact that the historical data amount of certain cables is insufficient in the recommendation process, the intelligent expansion network is used for expanding the data and training the model, a good recommendation effect is obtained, the recommendation accuracy is improved, and the recommendation problem of small-sample cable data is solved.
The foregoing description is only an overview of the technical solutions of the present invention, and in order to make the technical means of the present invention more clearly understood, the present invention may be implemented in accordance with the content of the description, and in order to make the above and other objects, features, and advantages of the present invention more clearly understood, the following preferred embodiments are described in detail with reference to the accompanying drawings.
Drawings
FIG. 1 is a flow chart of a cable transaction intelligent recommendation method according to an embodiment of the invention;
fig. 2 is a schematic diagram of a cable transaction intelligent recommendation system according to a second embodiment of the present invention.
Detailed Description
The present invention is further described below in conjunction with the following figures and specific examples so that those skilled in the art may better understand the present invention and practice it, but the examples are not intended to limit the present invention.
Example one
As shown in fig. 1, the cable transaction intelligent recommendation method in the preferred embodiment of the present invention includes the following steps:
s1, acquiring and sorting information of the cleaned cable, and establishing a commodity database;
the cable information includes, but is not limited to, one or more of materials, core number, current carrying capacity, insulation level, certification, application scenario, factory scale, factory location, and factory service level, and further includes insulation materials and sheaths, armor, bus type, and the like.
The sorting and cleaning comprises one or more of text error correction, keyword extraction, meaning analysis, feature coding and data standardization.
S2, acquiring basic information of the client, extracting historical behavior data of the client and establishing a client database;
wherein the customer base information includes, but is not limited to, one or more of name, gender, age, affiliated entity, position, and shipping address.
Wherein the historical behavior data includes, but is not limited to, one or more of search keywords, browsing records, purchasing records, communication records, collecting records, and rating records.
S3, judging whether the current client is a new client, if so, executing a step S4, otherwise, executing a step S8;
s4, calling the basic information of the new client in the client database, and obtaining the neighbor client through similarity comparison;
wherein, in the obtaining of the similarity between the neighboring clients by the similarity comparison, the similarity w between the two clientsuvComprises the following steps:
Figure BDA0003360174800000061
wherein the scoring vectors of the two clients u and v are respectively
Figure BDA0003360174800000062
And
Figure BDA0003360174800000063
alternatively, the similarity calculation method may be, but is not limited to, euclidean distance, cosine similarity, pearson correlation coefficient, hamming distance, and the like.
S5, deducing one or more application scenes of the new client application cable according to the transaction requirements of the neighbor clients;
the application scenario can be, but is not limited to, building, electric power, smelting, petrochemical, and electronic.
S6, calling cable parameters and manufacturer information in the commodity database according to the application scene to perform adaptability calculation to obtain adaptability scores corresponding to different cables;
wherein the suitability scores corresponding to the different cables are as follows:
Figure BDA0003360174800000064
where C is the nearest neighbor set of the new customer u, Rv,jScoring Cable j for neighboring customer v, Pu,jThe new customer is scored for the prediction of cable j.
S7, recommending one or more cables with the scores higher than those of the cables according to different application scenes, recording behavior feedback of a new client on a recommendation result, and updating a client database in real time;
wherein the behavior feedback may be, but is not limited to, click records, browsing records, collection records, purchase records, evaluation records.
S8, calling behavior data of the client in the client data, and calculating the degree of intention of the client to different cables through a collaborative filtering algorithm;
the collaborative filtering algorithm is a Model-based collaborative filtering algorithm, and the specific implementation algorithm can be, but is not limited to, an association algorithm, a clustering algorithm, a classification algorithm, a regression algorithm, a matrix decomposition algorithm, a neural network and a graph Model.
And S9, screening cables which may be interested by the customer according to the popularity, recording behavior data of the customer in the transaction process, and updating the customer database in real time.
Further, between the step S8 and the step S9, the method further includes the following steps:
A. and collecting the cable data with the least samples of the same type in the data, modeling and training by using an intelligent expansion network, obtaining a recommendation result with the highest similarity, and calculating the degree of intention of a client on the sample cable.
Specifically, step a includes:
a1, establishing two weight-sharing sub-networks with the same structure, inputting two groups of cable data X1And X2Respectively convert it into vector Gw(X1) And Gw(X2) Then, the distance E of the two vectors is calculated by a distance measurement methodw；
A2, assuming that the input sample of the intelligent expansion network is (X)1,X2Y), construct the loss function as follows:
Figure BDA0003360174800000071
L(W,(X1,X2,y)i)＝(1-y)LG(Ew(X1,X2)i)+yLI(Ew(X1,X2)i)
wherein (X)1,X2,y)iIs the ith input sample, where LGTo calculate the loss function for the cable for the same class only, LICalculating loss functions of different categories to the cable; the label y-0 indicates two sets of cable data X1And X2Of different types, the label y ═ 1 denotes two sets of cable data X1And X2Belonging to the same type, the setting of its loss function follows: when y is 0, the greater the distance, the smaller the loss, i.e. with respect to EwWhen y is 1, the smaller the distance, the greater the loss, i.e. with respect to EwIs a monotonically increasing function of.
A3, mixing LGSet to monotonically increase, set L to monotonically increaseIAnd setting the data to be monotonously reduced, and training a loss function by using the cable data with the least samples of the same type in the data.
Wherein, by separately designing the loss functions of the same class pair and the different class pairs, only L is needed to be minimized when the loss function is to be minimizedGDesigned to be monotonically increasing, and LIIs designed to be monotonically decreasing. The intelligent extension network is trained by utilizing the small sample cable data, so that a more accurate recommendation result is obtained.
The cable transaction intelligent recommendation method and system quantize the cable model parameters and the requirements of the customers on different scene applications into characteristics based on the recommendation algorithm of content and collaborative filtering, and convert manual research and big data screening into intelligent recommendation, so that the problem that the customers select cables with time and labor consumption is solved.
Meanwhile, the system considers the differentiation of new and old customers and solves the problem of cold start, and the system can stably and quickly recommend proper cables to all the customers, so that the sale of cable manufacturers is promoted, and the convenience of purchasing the cables by the customers is improved.
Secondly, aiming at the condition that the recommendation satisfaction degree is low due to the fact that the historical data amount of certain cables is insufficient in the recommendation process, the intelligent expansion network is used for expanding the data and training the model, a good recommendation effect is obtained, the recommendation accuracy is improved, and the recommendation problem of small-sample cable data is solved.
Example two
As shown in fig. 2, the present embodiment discloses a cable transaction intelligent recommendation system, which includes the following modules:
the commodity database establishing module is used for acquiring and arranging the cable cleaning information and establishing a commodity database;
the cable information includes, but is not limited to, one or more of materials, core number, current carrying capacity, insulation level, certification, application scenario, factory scale, factory location, and factory service level, and further includes insulation materials and sheaths, armor, bus type, and the like.
The sorting and cleaning comprises one or more of text error correction, keyword extraction, meaning analysis, feature coding and data standardization.
The client database establishing module is used for acquiring the basic information of the client, extracting the historical behavior data of the client and establishing a client database;
wherein the customer base information includes, but is not limited to, one or more of name, gender, age, affiliated entity, position, and shipping address.
Wherein the historical behavior data includes, but is not limited to, one or more of search keywords, browsing records, purchasing records, communication records, collecting records, and rating records.
The judging module is used for judging whether the current client is a new client or not, if so, the similarity comparing module is started, and if not, the popularity calculating module is started;
the similarity comparison module is used for calling the basic information of the new client in the client database and acquiring the neighbor client through similarity comparison;
optionally, in the obtaining of the similarity by the similarity comparison, the similarity w between the two clientsuvComprises the following steps:
Figure BDA0003360174800000091
wherein the scoring vectors of the two clients u and v are respectively
Figure BDA0003360174800000092
And
Figure BDA0003360174800000093
alternatively, the similarity calculation method may be, but is not limited to, euclidean distance, cosine similarity, pearson correlation coefficient, hamming distance, and the like.
The application scenario inference module is used for inferring one or more application scenarios of a new client application cable according to the transaction requirements of the neighbor clients;
the application scenario can be, but is not limited to, building, electric power, smelting, petrochemical, and electronic.
The adaptability calculation module is used for calling cable parameters and manufacturer information in the commodity database according to the application scene to perform adaptability calculation to obtain adaptability scores corresponding to different cables;
wherein the suitability scores corresponding to the different cables are as follows:
Figure BDA0003360174800000094
where C is the nearest neighbor set of the new customer u, Rv,jScoring Cable j for neighboring customer v, Pu,jThe new customer is scored for the prediction of cable j.
The cable recommendation module is used for recommending one or more cables with the grades earlier according to different application scenes, recording the behavior feedback of a new client on a recommendation result, and updating a client database in real time;
wherein the behavior feedback may be, but is not limited to, click records, browsing records, collection records, purchase records, evaluation records.
The system comprises a popularity calculation module, a collaborative filtering module and a data processing module, wherein the popularity calculation module is used for calling behavior data of a client in client data and calculating the popularity of the client to different cables through a collaborative filtering algorithm;
the collaborative filtering algorithm is a Model-based collaborative filtering algorithm, and the specific implementation algorithm can be, but is not limited to, an association algorithm, a clustering algorithm, a classification algorithm, a regression algorithm, a matrix decomposition algorithm, a neural network and a graph Model.
And the cable screening module is used for screening cables possibly interested by the customer according to the popularity, recording behavior data of the customer in the transaction process and updating the customer database in real time.
Further, the cable transaction intelligent recommendation system further comprises: and the small sample calculation module is used for collecting the cable data with the least number of samples of the same type in the data, modeling and training by using an intelligent expansion network, acquiring a recommendation result with the highest similarity, and calculating the degree of intention of a client on the sample cable.
The small sample calculation module is used for executing the following steps:
a1, establishing two weight-sharing sub-networks with the same structure, inputting two groups of cable data X1And X2Respectively convert it into vector Gw(X1) And Gw(X2) Then, the distance E of the two vectors is calculated by a distance measurement methodw；
A2, assuming that the input sample of the intelligent expansion network is (X)1,X2Y), construct the loss function as follows:
Figure BDA0003360174800000101
L(W,(X1,X2,y)i)＝(1-y)LG(Ew(X1,X2)i)+yLI(Ew(X1,X2)i)
wherein (X)1,X2,y)iIs the ith input sample, where LGTo calculate the loss function for the cable for the same class only, LICalculating loss functions of different categories to the cable; the label y-0 indicates two sets of cable data X1And X2Of different types, the label y ═ 1 denotes two sets of cable data X1And X2Belonging to the same type, the setting of its loss function follows: when y is 0, the greater the distance, the smaller the loss, i.e. with respect to EwWhen y is 1, the smaller the distance, the greater the loss, i.e. with respect to EwIs a monotonically increasing function of.
A3, mixing LGSet to monotonically increase, set L to monotonically increaseIAnd setting the data to be monotonously reduced, and training a loss function by using the cable data with the least samples of the same type in the data.
Wherein, by separately designing the loss functions of the same class pair and the different class pairs, only L is needed to be minimized when the loss function is to be minimizedGDesigned to be monotonically increasing, and LIIs designed to be monotonically decreasing. The intelligent extension network is trained by utilizing the small sample cable data, so that a more accurate recommendation result is obtained.
Further, the cable transaction intelligent recommendation system further comprises: a database and a display module.
And the database is used for storing relevant data of the cable and the client.
And the display module is used for interacting with the client, displaying the intelligent recommendation result of the cable transaction recommendation system and transmitting the behavior data of the client back to the cable transaction recommendation system.
The above embodiments are merely preferred embodiments for fully illustrating the present invention, and the scope of the present invention is not limited thereto. The equivalent substitution or change made by the technical personnel in the technical field on the basis of the invention is all within the protection scope of the invention. The protection scope of the invention is subject to the claims.

Claims (10)

1. An intelligent cable transaction recommendation method is characterized by comprising the following steps:
s1, acquiring and sorting information of the cleaned cable, and establishing a commodity database;
s2, acquiring basic information of the client, extracting historical behavior data of the client and establishing a client database;
s3, judging whether the current client is a new client, if so, executing a step S4, otherwise, executing a step S8;
s4, calling the basic information of the new client in the client database, and obtaining the neighbor client through similarity comparison;
s5, deducing one or more application scenes of the new client application cable according to the transaction requirements of the neighbor clients;
s6, calling cable parameters and manufacturer information in the commodity database according to the application scene to perform adaptability calculation to obtain adaptability scores corresponding to different cables;
s7, recommending one or more cables with the scores higher than those of the cables according to different application scenes, recording behavior feedback of a new client on a recommendation result, and updating a client database in real time;
s8, calling behavior data of the client in the client data, and calculating the degree of intention of the client to different cables through a collaborative filtering algorithm;
and S9, screening cables which may be interested by the customer according to the popularity, recording behavior data of the customer in the transaction process, and updating the customer database in real time.
2. The intelligent cable transaction recommendation method of claim 1, wherein between the step S8 and the step S9, further comprising the steps of:
A. and collecting the cable data with the least samples of the same type in the data, modeling and training by using an intelligent expansion network, obtaining a recommendation result with the highest similarity, and calculating the degree of intention of a client on the sample cable.
3. The intelligent cable transaction recommendation method of claim 2, wherein step a comprises:
a1, establishing two weight-sharing sub-networks with the same structure, inputting two groups of cable data X1And X2Respectively convert it into vector Gw(X1) And Gw(X2) Then, the distance E of the two vectors is calculated by a distance measurement methodw；
A2, assuming that the input sample of the intelligent expansion network is (X)1,X2Y), construct the loss function as follows:
Figure FDA0003360174790000021
L(W,(X1,X2,y)i)＝(1-y)LG(Ew(X1,X2)i)+yLI(Ew(X1,X2)i)
wherein (X)1,X2,y)iIs the ith input sample, where LGTo calculate the loss function for the cable for the same class only, LICalculating loss functions of different categories to the cable; the label y-0 indicates two sets of cable data X1And X2Of different types, the label y ═ 1 denotes two sets of cable data X1And X2Belonging to the same type, the setting of its loss function follows: when y is 0, the greater the distance, the smaller the loss, i.e. with respect to EwWhen y is 1, the smaller the distance, the greater the loss, i.e. with respect to EwIs a monotonically increasing function of.
A3, mixing LGSet to monotonically increase, set L to monotonically increaseIAnd setting the data to be monotonously reduced, and training a loss function by using the cable data with the least samples of the same type in the data.
4. As claimed inThe intelligent cable transaction recommendation method of claim 1 is characterized in that in the similarity comparison to obtain the neighbor clients, the similarity w between two clientsuvComprises the following steps:
Figure FDA0003360174790000022
wherein the scoring vectors of the two clients u and v are respectively
Figure FDA0003360174790000023
And
Figure FDA0003360174790000024
5. the intelligent recommendation method for cable transaction as claimed in claim 1, wherein the suitability scores corresponding to different cables are as follows:
Figure FDA0003360174790000025
where C is the nearest neighbor set of the new customer u, Rv,jScoring Cable j for neighboring customer v, Pu,jThe new customer is scored for the prediction of cable j.
6. The method for intelligently recommending cable transactions according to claim 1, wherein said cable information includes one or more of materials, core count, ampacity, insulation level, certification, application scenario, factory size, factory location, and factory service level, and further includes insulation and jacket, armor, and bus type.
7. The cable transaction intelligent recommendation method of claim 1, wherein the customer base information comprises one or more of name, gender, age, affiliated entity, position and shipping address.
8. The cable transaction intelligent recommendation method of claim 1, wherein the historical behavior data comprises one or more of search keywords, browsing records, purchasing records, communication records, collecting records, and rating records.
9. The intelligent cable transaction recommendation method of claim 1, wherein the cleaning comprises one or more of text correction, keyword extraction, lexical analysis, feature coding, and data normalization.
10. A cable transaction intelligent recommendation system, comprising:
the commodity database establishing module is used for acquiring and arranging the cable cleaning information and establishing a commodity database;
the client database establishing module is used for acquiring the basic information of the client, extracting the historical behavior data of the client and establishing a client database;
the judging module is used for judging whether the current client is a new client or not;
the similarity comparison module is used for calling the basic information of the new client in the client database and acquiring the neighbor client through similarity comparison;
the application scenario inference module is used for inferring one or more application scenarios of a new client application cable according to the transaction requirements of the neighbor clients;
the adaptability calculation module is used for calling cable parameters and manufacturer information in the commodity database according to the application scene to perform adaptability calculation to obtain adaptability scores corresponding to different cables;
the cable recommendation module is used for recommending one or more cables with the grades earlier according to different application scenes, recording the behavior feedback of a new client on a recommendation result, and updating a client database in real time;
the system comprises a popularity calculation module, a collaborative filtering module and a data processing module, wherein the popularity calculation module is used for calling behavior data of a client in client data and calculating the popularity of the client to different cables through a collaborative filtering algorithm;
and the cable screening module is used for screening cables possibly interested by the customer according to the popularity, recording behavior data of the customer in the transaction process and updating the customer database in real time.
CN202111364667.6A 2021-11-17 2021-11-17 Cable transaction intelligent recommendation method and system Active CN114066572B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111364667.6A CN114066572B (en) 2021-11-17 2021-11-17 Cable transaction intelligent recommendation method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111364667.6A CN114066572B (en) 2021-11-17 2021-11-17 Cable transaction intelligent recommendation method and system

Publications (2)

Publication Number Publication Date
CN114066572A true CN114066572A (en) 2022-02-18
CN114066572B CN114066572B (en) 2022-07-12

Family

ID=80277602

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111364667.6A Active CN114066572B (en) 2021-11-17 2021-11-17 Cable transaction intelligent recommendation method and system

Country Status (1)

Country Link
CN (1) CN114066572B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060041548A1 (en) * 2004-07-23 2006-02-23 Jeffrey Parsons System and method for estimating user ratings from user behavior and providing recommendations
CN106471491A (en) * 2015-05-29 2017-03-01 深圳市汇游智慧旅游网络有限公司 A kind of collaborative filtering recommending method of time-varying
CN109684538A (en) * 2018-12-03 2019-04-26 重庆邮电大学 A kind of recommended method and recommender system based on individual subscriber feature
CN110083764A (en) * 2019-04-11 2019-08-02 东华大学 A kind of collaborative filtering cold start-up way to solve the problem
CN110084610A (en) * 2019-04-23 2019-08-02 东华大学 A kind of network trading fraud detection system based on twin neural network
CN110287402A (en) * 2019-05-05 2019-09-27 江苏一乙生态农业科技有限公司 A kind of personalized reviews recommended method based on user preference
CN110619552A (en) * 2018-06-19 2019-12-27 航天信息股份有限公司 Member shopping data mining algorithm comprehensive engine
CN111899075A (en) * 2020-08-11 2020-11-06 恒瑞通（福建）信息技术有限公司 Personalized commodity recommendation method and device based on user behaviors
CN112581189A (en) * 2020-12-29 2021-03-30 科技谷（厦门）信息技术有限公司 Intelligent supplier recommendation system and method
CN113192626A (en) * 2021-04-13 2021-07-30 山东大学 Medicine taking scheme recommendation system and method based on twin neural network

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060041548A1 (en) * 2004-07-23 2006-02-23 Jeffrey Parsons System and method for estimating user ratings from user behavior and providing recommendations
CN106471491A (en) * 2015-05-29 2017-03-01 深圳市汇游智慧旅游网络有限公司 A kind of collaborative filtering recommending method of time-varying
CN110619552A (en) * 2018-06-19 2019-12-27 航天信息股份有限公司 Member shopping data mining algorithm comprehensive engine
CN109684538A (en) * 2018-12-03 2019-04-26 重庆邮电大学 A kind of recommended method and recommender system based on individual subscriber feature
CN110083764A (en) * 2019-04-11 2019-08-02 东华大学 A kind of collaborative filtering cold start-up way to solve the problem
CN110084610A (en) * 2019-04-23 2019-08-02 东华大学 A kind of network trading fraud detection system based on twin neural network
CN110287402A (en) * 2019-05-05 2019-09-27 江苏一乙生态农业科技有限公司 A kind of personalized reviews recommended method based on user preference
CN111899075A (en) * 2020-08-11 2020-11-06 恒瑞通（福建）信息技术有限公司 Personalized commodity recommendation method and device based on user behaviors
CN112581189A (en) * 2020-12-29 2021-03-30 科技谷（厦门）信息技术有限公司 Intelligent supplier recommendation system and method
CN113192626A (en) * 2021-04-13 2021-07-30 山东大学 Medicine taking scheme recommendation system and method based on twin neural network

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
DONG CAO 等: "Alleviating the New Item Cold-Start Problem by Combining Image Similarity", 《2019 IEEE 9TH INTERNATIONAL CONFERENCE ON ELECTRONICS INFORMATION AND EMERGENCY COMMUNICATION》 *
JIAN DONG 等: "Cooperative Filtering Program Recommendation Algorithm Based on User Situations and Missing Values Estimation", 《INTERNATIONAL CONFERENCE ON INTELLIGENT INFORMATION PROCESSING》 *
李昊: "基于微信小程序的智能推荐点餐系统的设计与实现", 《中国优秀硕士学位论文全文数据库信息科技辑》 *
蔡诚 等: "分系统级电缆网接点智能匹配方法", 《2020中国自动化大会(CAC2020)论文集》 *

Also Published As

Publication number Publication date
CN114066572B (en) 2022-07-12

Similar Documents

Publication Publication Date Title
US20070118546A1 (en) User's preference prediction from collective rating data
CN106708821A (en) User personalized shopping behavior-based commodity recommendation method
CN111523055B (en) Collaborative recommendation method and system based on agricultural product characteristic attribute comment tendency
CN101990668A (en) Recommendation information generation apparatus and recommendation information generation method
Sriram et al. A review on multi-criteria decision-making and its application
CN109816015B (en) Recommendation method and system based on material data
CN113239319A (en) Method for automatically matching and pushing supplier to bid and quote
CN114266443A (en) Data evaluation method and device, electronic equipment and storage medium
CN116187808A (en) Electric power package recommendation method based on virtual power plant user-package label portrait
CN115631012A (en) Target recommendation method and device
Hassan et al. Improving prediction accuracy of multi-criteria recommender systems using adaptive genetic algorithms
CN114066572B (en) Cable transaction intelligent recommendation method and system
CN112860878A (en) Service data recommendation method, storage medium and equipment
CN116663909A (en) Provider risk identification data processing method and device
CN113450167A (en) Commodity recommendation method and device
CN114429384B (en) Intelligent product recommendation method and system based on e-commerce platform
CN116186541A (en) Training method and device for recommendation model
CN110879821A (en) Method, device, equipment and storage medium for generating rating card model derivative label
CN115829683A (en) Power integration commodity recommendation method and system based on inverse reward learning optimization
CN113837843B (en) Product recommendation method and device, medium and electronic equipment
CN115907926A (en) Commodity recommendation method and device, electronic equipment and storage medium
CN112200602B (en) Neural network model training method and device for advertisement recommendation
CN110956528B (en) Recommendation method and system for e-commerce platform
CN113094584A (en) Method and device for determining recommended learning resources
CN113761002A (en) Information pushing method, device, equipment and computer readable storage medium

Legal Events

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