CN114155057A - Commodity recommendation system for electronic commerce platform - Google Patents
Commodity recommendation system for electronic commerce platform Download PDFInfo
- Publication number
- CN114155057A CN114155057A CN202111386092.8A CN202111386092A CN114155057A CN 114155057 A CN114155057 A CN 114155057A CN 202111386092 A CN202111386092 A CN 202111386092A CN 114155057 A CN114155057 A CN 114155057A
- Authority
- CN
- China
- Prior art keywords
- commodity
- user
- module
- recommended
- search
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Abstract
The invention relates to a recommendation system, in particular to a commodity recommendation system for an e-commerce platform, which comprises a server, a search content acquisition module and a basic information acquisition module, wherein after the basic information acquisition module acquires relevant basic information of a user to be recommended, the server extracts interest tag vectors of the user to be recommended through a basic information analysis module and classifies the user to be recommended through a recommended user classification module, the server constructs a corresponding relation between a commodity category and a user category through a classification relation construction module and outputs corresponding recommended commodities to the user to be recommended through a recommended commodity output module; the technical scheme provided by the invention can effectively overcome the defect that the electronic commerce platform in the prior art cannot provide accurate commodity recommendation service for the user.
Description
Technical Field
The invention relates to a recommendation system, in particular to a commodity recommendation system for an electronic commerce platform.
Background
The electronic commerce platform is a platform for providing online transaction negotiation service for enterprises or individuals, and is a virtual network space for carrying out business activities on the Internet and a virtual environment for ensuring smooth operation of the businesses; the system is an important medium for coordinating and integrating information flow, cargo flow and fund flow in order, relevance and high-efficiency flow. Enterprises and merchants can fully utilize shared resources such as network infrastructure, payment platforms, security platforms, management platforms and the like provided by the electronic commerce platform, and can effectively develop own commercial activities at low cost.
Most of the existing E-commerce platforms have a commodity recommendation function, and two methods for judging user preference are available, wherein one method is to utilize the scoring information of a user on commodities, search neighbors with similar scoring behaviors based on a collaborative filtering method, and recommend commodities liked by the neighbors to the user; and secondly, recommending interested commodities for the user by using the registration information and the commodity basic information of the user through a content recommendation method. At present, each website is similar in recommendation method, namely, a data source related to commodities is analyzed firstly, and the preference degree of a user to each commodity is calculated. When commodities need to be recommended to a user, a preference matrix of the user for the commodities is obtained, the recommended number is given, and the commodities with the highest preference values are recommended to the user.
The existing commodity recommendation system has the following defects: the preference mining between the user and the commodities is not sufficient, so that a large gap exists between the commodities recommended to the user at last and the commodities really liked by the user, the user is prone to dislike, the recommendation is performed only based on the scoring relation between the user and the commodities, and other important reference data sources are omitted.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects in the prior art, the invention provides a commodity recommendation system for an e-commerce platform, which can effectively overcome the defect that the e-commerce platform in the prior art cannot provide accurate commodity recommendation service for users.
(II) technical scheme
In order to achieve the purpose, the invention is realized by the following technical scheme:
a commodity recommendation system for an electronic commerce platform comprises a server, a search content acquisition module and a basic information acquisition module, wherein the server constructs a commodity search model for searching corresponding commodities in a commodity database according to word characteristics through a search model construction module, trains and optimizes the commodity search model through a search model training module and a search model optimization module respectively, after the search content acquisition module acquires search contents input by a user in a search box, the server extracts keywords from the search contents through a keyword extraction module and extracts the word characteristics from the keywords through a word characteristic extraction module, and the server optimally displays the search results of the commodity search model through an optimization display module according to the word characteristics;
after the basic information acquisition module acquires the relevant basic information of the user to be recommended, the server extracts the interest tag vector of the user to be recommended through the basic information analysis module and classifies the user to be recommended through the recommended user classification module, the server establishes the corresponding relation between the commodity category and the user category through the classification relation establishment module, and outputs the corresponding recommended commodity to the user to be recommended through the recommended commodity output module.
Preferably, the commodity search model constructed by the search model construction module is a neural network model, the search model training module calls pictures of each commodity from the commodity database, selects a proper area on the pictures according to the commodity information editing content of the commodity, marks corresponding keywords, and inputs the pictures marked with the keywords into the commodity search model for training.
Preferably, after the search model training module trains the commodity search model for a period of time, the search model optimization module evaluates the accuracy of the search result of the commodity search model in the corresponding period of time, and optimizes the commodity search model according to the evaluation result by a random gradient descent method.
Preferably, the keyword extraction module extracts keywords from the search content through natural language analysis, and the word feature extraction module extracts at least one word feature from the keywords and performs normalization processing on the word features.
Preferably, the term feature extraction module inputs the extracted term features into the optimized commodity search model, and the optimization display module optimizes and displays the search results of the commodity search model by integrating the consumption level of the user, frequent consumption stores and sales volume of each search commodity.
Preferably, the basic information analysis module preprocesses the basic information related to the user to be recommended acquired by the basic information acquisition module, and extracts the interest tag vector and the corresponding vector weight of the user to be recommended from the preprocessed related basic information.
Preferably, the recommending user classifying module performs weighted fusion on the interest tag vectors to obtain the interest features of the users to be recommended, and classifies the users to be recommended through a clustering algorithm based on the interest features.
Preferably, the recommending user classifying module classifies users to be recommended through a clustering algorithm based on the interest features, and includes:
s1, randomly extracting k interest features of each user from the user database, and respectively using the k interest features as category centers of k user categories;
s2, calculating the similarity of the interest characteristics of other users to the center of each category, and classifying the users into the user category with the highest similarity;
s3, recalculating the category centers of the user categories, and performing clustering operation again;
s4, repeating S2 and S3 until the dissimilarity between the clustering result of the current round and the clustering result of the previous round is smaller than a set threshold;
and S5, taking the interest characteristics of the user to be recommended and the user category corresponding to the nearest category center in the S4 as the user category of the user to be recommended.
Preferably, the classification relationship building module builds a corresponding relationship between the commodity category and the user category, and includes:
acquiring commodity data corresponding to each commodity in a commodity database, and constructing a commodity label vector;
dividing the commodities into different commodity categories according to the commodity label vectors, and calculating the matching degree between the commodity categories and the user categories;
and constructing a corresponding relation between the commodity category and the user category based on the matching degree.
(III) advantageous effects
Compared with the prior art, the commodity recommendation system for the e-commerce platform provided by the invention has the advantages that the matching degree between the search result and the search requirement of the user can be effectively improved by establishing the commodity search model and optimizing the commodity search model; after the relevant basic information of the user to be recommended is obtained, the interest tag vector of the user to be recommended is extracted, the user to be recommended is classified based on the interest tag vector, corresponding recommended commodities are output to the user to be recommended by constructing the corresponding relation between the commodity category and the user category, and therefore accurate commodity recommendation service can be provided for the user.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
FIG. 1 is a schematic diagram of the system of the present invention;
fig. 2 is a schematic flow chart illustrating a process of outputting a corresponding recommended commodity to a user to be recommended by using a recommended commodity output module according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
A commodity recommendation system for an electronic commerce platform is disclosed, as shown in figure 1, and comprises a server, a search content acquisition module and a basic information acquisition module, wherein the server constructs a commodity search model for searching corresponding commodities in a commodity database according to word characteristics through a search model construction module, trains and optimizes the commodity search model through a search model training module and a search model optimization module respectively, after the search content acquisition module acquires search contents input by a user in a search box, the server extracts keywords through a keyword extraction module and word characteristics through a word characteristic extraction module, and the server optimally displays the search results of the commodity search model through an optimization display module according to the word characteristics.
The commodity search model constructed by the search model construction module is a neural network model, the search model training module calls pictures of commodities from the commodity database, selects a proper area on the pictures according to commodity information editing contents of the commodities and marks corresponding keywords, and inputs the pictures marked with the keywords into the commodity search model for training.
After the commodity search model is trained for a period of time by the search model training module, the accuracy of the search result of the commodity search model in the corresponding period of time is evaluated by the search model optimizing module, and the commodity search model is optimized by a random gradient descent method according to the evaluation result. The commodity search model is optimized, and the matching degree between the search result and the search requirement of the user can be effectively improved.
The keyword extraction module extracts keywords from the search content through natural language analysis, and the word feature extraction module extracts at least one word feature from the keywords and performs normalization processing on the word features. The term features processed through normalization have the same length, so that the optimized commodity search model can search commodities quickly according to the term features of the keywords.
The word characteristic extraction module inputs the extracted word characteristics into the optimized commodity search model, and the optimization display module integrates the consumption level of the user, frequent consumption shops and the sales volume of each searched commodity to optimize and display the search results of the commodity search model.
When the optimized display module is used for optimally displaying the search results of the commodity search model, the search results of the commodity search model are optimally sorted according to the priority sequence of frequent shopping malls, the sales volume of the searched commodities and the consumption level, and are displayed in the lower area of the input search box of the user.
As shown in fig. 1 and 2, after the basic information obtaining module obtains the relevant basic information of the user to be recommended, the server extracts the interest tag vector of the user to be recommended through the basic information analyzing module and classifies the user to be recommended through the recommended user classifying module, the server constructs the corresponding relationship between the commodity category and the user category through the classification relationship constructing module, and outputs the corresponding recommended commodity to the user to be recommended through the recommended commodity output module.
The basic information analysis module is used for preprocessing the basic information (including personal registration information, real-time browsing record, real-time comment data and the like during user registration) of the user to be recommended acquired by the basic information acquisition module, and extracting the interest tag vector and the corresponding vector weight of the user to be recommended from the preprocessed basic information.
The recommendation user classification module performs weighted fusion on the interest tag vectors (by using each interest tag vector and corresponding vector weight) to obtain the interest features of the users to be recommended, and classifies the users to be recommended through a clustering algorithm based on the interest features.
The recommendation user classification module classifies users to be recommended through a clustering algorithm based on the interest characteristics, and comprises the following steps:
s1, randomly extracting k interest features of each user from the user database, and respectively using the k interest features as category centers of k user categories;
s2, calculating the similarity of the interest characteristics of other users to the center of each category, and classifying the users into the user category with the highest similarity;
s3, recalculating the category centers of the user categories, and performing clustering operation again;
s4, repeating S2 and S3 until the dissimilarity between the clustering result of the current round and the clustering result of the previous round is smaller than a set threshold;
and S5, taking the interest characteristics of the user to be recommended and the user category corresponding to the nearest category center in the S4 as the user category of the user to be recommended.
The classification relation building module builds the corresponding relation between the commodity category and the user category, and comprises the following steps:
acquiring commodity data corresponding to each commodity in a commodity database, and constructing a commodity label vector;
dividing the commodities into different commodity categories according to the commodity label vectors, and calculating the matching degree between the commodity categories and the user categories;
and constructing a corresponding relation between the commodity category and the user category based on the matching degree.
According to the technical scheme, the users to be recommended are classified through a clustering algorithm based on the interest characteristics, so that the users to be recommended can be accurately classified according to the related basic information of the users to be recommended, the corresponding relation between the commodity category and the user category is accurately established according to the matching degree between the commodity category and the user category obtained through calculation, accurate association between the commodity category and the user category is further achieved, and the electronic commerce platform can provide accurate commodity recommendation service for the users.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (9)
1. A commodity recommendation system for an electronic commerce platform is characterized in that: the system comprises a server, a search content acquisition module and a basic information acquisition module, wherein the server constructs a commodity search model for searching corresponding commodities in a commodity database according to word characteristics through a search model construction module, trains and optimizes the commodity search model through a search model training module and a search model optimization module respectively, after the search content acquisition module acquires search contents input by a user in a search box, the server extracts keywords from the search contents through a keyword extraction module and extracts the word characteristics from the keywords through a word characteristic extraction module, and the server optimally displays the search results of the commodity search model through an optimization display module according to the word characteristics;
after the basic information acquisition module acquires the relevant basic information of the user to be recommended, the server extracts the interest tag vector of the user to be recommended through the basic information analysis module and classifies the user to be recommended through the recommended user classification module, the server establishes the corresponding relation between the commodity category and the user category through the classification relation establishment module, and outputs the corresponding recommended commodity to the user to be recommended through the recommended commodity output module.
2. The merchandise recommendation system for an e-commerce platform of claim 1, wherein: the commodity search model constructed by the search model construction module is a neural network model, the search model training module calls pictures of commodities from a commodity database, selects a proper area on the pictures according to commodity information editing contents of the commodities and marks corresponding keywords, and inputs the pictures marked with the keywords into the commodity search model for training.
3. The merchandise recommendation system for an e-commerce platform of claim 2, wherein: after the search model training module trains the commodity search model for a period of time, the search model optimization module evaluates the accuracy of the search result of the commodity search model in the corresponding period of time, and optimizes the commodity search model by a random gradient descent method according to the evaluation result.
4. The merchandise recommendation system for an e-commerce platform of claim 1, wherein: the keyword extraction module extracts keywords from search contents through natural language analysis, and the word feature extraction module extracts at least one word feature from the keywords and performs normalization processing on the word features.
5. The merchandise recommendation system for an e-commerce platform of claim 4, wherein: the term feature extraction module inputs the extracted term features into the optimized commodity search model, and the optimization display module integrates the consumption level of the user, frequent consumption shops and the sales volume of each searched commodity to optimize and display the search results of the commodity search model.
6. The merchandise recommendation system for an e-commerce platform of claim 1, wherein: the basic information analysis module preprocesses the basic information of the user to be recommended acquired by the basic information acquisition module, and extracts the interest tag vector and the corresponding vector weight of the user to be recommended from the preprocessed basic information.
7. The merchandise recommendation system for an e-commerce platform of claim 6, wherein: the recommendation user classification module performs weighted fusion on the interest label vectors to obtain the interest characteristics of the users to be recommended, and classifies the users to be recommended through a clustering algorithm based on the interest characteristics.
8. The merchandise recommendation system for an e-commerce platform of claim 7, wherein: the recommendation user classification module classifies users to be recommended through a clustering algorithm based on interest characteristics, and comprises the following steps:
s1, randomly extracting k interest features of each user from the user database, and respectively using the k interest features as category centers of k user categories;
s2, calculating the similarity of the interest characteristics of other users to the center of each category, and classifying the users into the user category with the highest similarity;
s3, recalculating the category centers of the user categories, and performing clustering operation again;
s4, repeating S2 and S3 until the dissimilarity between the clustering result of the current round and the clustering result of the previous round is smaller than a set threshold;
and S5, taking the interest characteristics of the user to be recommended and the user category corresponding to the nearest category center in the S4 as the user category of the user to be recommended.
9. The merchandise recommendation system for an e-commerce platform of claim 7, wherein: the classification relation building module builds the corresponding relation between the commodity category and the user category, and comprises the following steps:
acquiring commodity data corresponding to each commodity in a commodity database, and constructing a commodity label vector;
dividing the commodities into different commodity categories according to the commodity label vectors, and calculating the matching degree between the commodity categories and the user categories;
and constructing a corresponding relation between the commodity category and the user category based on the matching degree.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111386092.8A CN114155057A (en) | 2021-11-22 | 2021-11-22 | Commodity recommendation system for electronic commerce platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111386092.8A CN114155057A (en) | 2021-11-22 | 2021-11-22 | Commodity recommendation system for electronic commerce platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114155057A true CN114155057A (en) | 2022-03-08 |
Family
ID=80457319
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111386092.8A Pending CN114155057A (en) | 2021-11-22 | 2021-11-22 | Commodity recommendation system for electronic commerce platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114155057A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115796924A (en) * | 2023-01-31 | 2023-03-14 | 武汉亿诚同创科技有限公司 | Cloud platform e-commerce data processing method and system based on big data |
-
2021
- 2021-11-22 CN CN202111386092.8A patent/CN114155057A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115796924A (en) * | 2023-01-31 | 2023-03-14 | 武汉亿诚同创科技有限公司 | Cloud platform e-commerce data processing method and system based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110222272B (en) | Potential customer mining and recommending method | |
| CN107861972B (en) | Method and equipment for displaying full commodity result after user inputs commodity information | |
| CN107833082B (en) | Commodity picture recommendation method and device | |
| CN107391493B (en) | Public opinion information extraction method and device, terminal equipment and storage medium | |
| CN109711955B (en) | Poor evaluation early warning method and system based on current order and blacklist base establishment method | |
| KR102142126B1 (en) | Hierarchical Category Cluster Based Shopping Basket Associated Recommendation Method | |
| CN111667337A (en) | Commodity evaluation ordering method and system | |
| CN111047412A (en) | Big data electricity merchant operation platform | |
| CN113946754A (en) | User portrait based rights and interests recommendation method, device, equipment and storage medium | |
| CN114266443A (en) | Data evaluation method and device, electronic equipment and storage medium | |
| CN114596031A (en) | Express terminal user portrait model based on full life cycle data | |
| CN114997956A (en) | Mother and infant product intelligent recommendation system based on big data | |
| CN111429161A (en) | Feature extraction method, feature extraction device, storage medium, and electronic apparatus | |
| CN113837844A (en) | Multi-cascade downstream enterprise recommendation system and method and storage medium | |
| CN114155057A (en) | Commodity recommendation system for electronic commerce platform | |
| CN112949322A (en) | E-commerce opinion mining recommendation system driven by online text comments | |
| KR102404247B1 (en) | Customer management system | |
| KR101754124B1 (en) | The restaurant recommending system and the recommending method thereof | |
| CN115563176A (en) | Electronic commerce data processing system and method | |
| US20240062264A1 (en) | Ai- backed e-commerce for all the top rated products on a single platform | |
| CN115131108A (en) | E-commerce commodity screening system | |
| CN113254775A (en) | Credit card product recommendation method based on client browsing behavior sequence | |
| CN114169926A (en) | Commodity data analysis method, system, device and medium based on user comments | |
| KR20180092053A (en) | Intelligent shopping managing system | |
| CN113822691A (en) | User account identification method, device, system and 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 |