CN102831543A - E-commerce recommendation method - Google Patents
E-commerce recommendation method Download PDFInfo
- Publication number
- CN102831543A CN102831543A CN2012103485880A CN201210348588A CN102831543A CN 102831543 A CN102831543 A CN 102831543A CN 2012103485880 A CN2012103485880 A CN 2012103485880A CN 201210348588 A CN201210348588 A CN 201210348588A CN 102831543 A CN102831543 A CN 102831543A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- recommend method
- recommendation
- relevant
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an e-commerce recommendation method which comprises an active recommendation method and a passive recommendation method, wherein the active recommendation method comprises the following steps of: I, when in user access, selecting a subscription type by a user, setting a subscription keyword, and selecting the related industry; II, storing the subscription conditions set by the user in a database; III, extracting the related fields corresponding to the subscription conditions, and performing acquaintance matching and sorting according to the related fields; and IV, recommending the processed information to the user. The method disclosed by the invention introduces the e-commerce recommendation system idea, analyzes the access behaviors (such as access, search, evaluation, collection and the like) of a customer on the e-commerce website by use of the technologies such as statistics, artificial intelligence data mining and the like, and generates a recommendation result of the business opportunity information favored by the customer so as to help the customer accurately catch the business opportunity in time.
Description
Technical field
The present invention relates to a kind of e-commerce services platform, be specifically related to a kind of ecommerce recommend method.
Background technology
Along with Development of E-business, the commodity amount in the e-commerce platform is exponential increase, has formed mass data, and the client is difficult to from the commodity ocean of vastness like this, pick out his valuable business opportunity.How business opportunity information is effectively organized and showed how to understand client's interest and hobby as much as possible, thereby make things convenient for the user to obtain business opportunity to become the problem that presses for solution in the e-commerce development process to optimize website design.
Summary of the invention
In view of this; The purpose of this invention is to provide a kind of ecommerce recommend method; Introduce the Technologies of Recommendation System in E-Commerce theory, utilize technology such as statistics, artificial intelligence data mining, analyze the visit behavior (like visit, search, comment, collection etc.) of client at e-commerce website; Produce the recommendation results of the interested business opportunity information of client, thereby help the client to capture business opportunity timely and accurately.
To achieve these goals, the present invention adopts following technical scheme:
A kind of ecommerce recommend method wherein, comprises active recommend method and passive type recommend method, and said active recommend method comprises the steps:
Step 1 during user capture, selects to subscribe to type by the user, sets and subscribes to keyword, selects relevant industries;
Step 2, the subscription condition that the user is set stores database into;
Step 3 is extracted the corresponding relevant field of subscription condition, according to relevant field acquaintance property coupling and ordering;
Step 4 is with the user that recommends after handling.
As preferably, said passive type recommend method comprises the steps:
Step 1, when user capture, all operations behavior and related data information in the recording user access process;
Step 2 is put in order user data, forms data warehouse;
Step 3 is analyzed the data in the data warehouse, and extracts representative content characteristic values;
Step 4 is formulated recommendation rules, and sets relevant data field;
Step 5 is carried out correlation analysis and knowledge and magnanimity matching operation mutually according to recommendation rules to relevant data content eigenwert;
Step 6 the weight information of the data after the coupling according to content characteristic values sorted, and the data after will sorting is recommended to show the client.
As preferably, the user data in the said passive type recommend method step 2 comprises supply data, businessman's data and prosperous shop click data.
Beneficial effect of the present invention is:
The present invention includes active recommendation and passive type and recommend two kinds of ways of recommendation.Active recommendation is meant that system according to the analysis to user profile and behavior, provides user's interest business opportunity information.And passive type recommends to be meant that the effort that the user passes through oneself obtains needed business opportunity information under system help.The present invention recommends that active recommendation and passive type combine, to reach best recommendation effect.
Adopt the lucence technology to combine information filtering and two kinds of algorithms of collaborative filtering to realize recommending on the recommend method of the present invention.The advantage of this recommend method is:
Effective integration the search and recommend integrated.Search in Website engine of the present invention and recommended engine all adopt the Lucene technology, are that search and recommendation effectively integrate.Search is the most frequent, most important user's operation behavior.It is integrated that commending system can be obtained is more timely, more effective, more comprehensive user data, and then recommendation service more accurately is provided.
2. adopt the lucene technology to carry out the phase knowledge and magnanimity analysis of data, weight sorting operation.Knowledge and magnanimity analysis of data phase and weight ordering are the core technologies in the commending system, and he is directly connected to the accuracy of recommending user data.At the text participle, the phase knowledge and magnanimity are analyzed, aspects such as full-text search, and Lucene compares with other technologies and is in the low level that holds a safe lead.Therefore Lucene has powerful data acquaintance property analysis and weight ranking function.Compare the Lucene technology combines information filtering and collaborative filtering to provide more accurately more efficiently to wait recommendation service with suggested design commonly used.
3. the present invention adopts according to different recommended requirements and selects different recommend methods, will adopt information filtering and collaborative filtering to combine simultaneously.The information filtering algorithm is based on merchandise news, comprises that the attribute of commodity and correlativity between the commodity and client's happiness dislike to its recommendation.Its drawback is and can not can only finds the interesting similar commodity with the user for the user finds new interested commodity, can not find new interested commodity for the user.The behavior that collaborative filtering basis and active user have a user of similar viewpoint is recommended this user and is predicted.Collaborative filtering does not need the description of commodity characteristic, and what its was learnt is the similarity between user's buying behavior, and not commodity-dependent characteristic.
Description of drawings
Fig. 1 is the process flow diagram of the active recommendation of the present invention;
The process flow diagram that Fig. 2 recommends for passive type of the present invention.
Embodiment
Through accompanying drawing the present invention is done further description below:
The effect that the ecommerce recommend method plays in e-commerce website mainly comprises: (1) can be in the identification client's who analyzes the visit behavior of visitor in e-commerce website hobby and demand; And to lead referral to its useful business opportunity information (such as commodity; Buying; The purchaser, supplier etc.).(2) attract more visitors, increase the website visiting amount.(3) exposure of raising particular commodity increases the addressing machine meeting of commodity.(4) increase the client in the prosperous residence time of spreading, browse more commodity.(5) improve the conversion ratio that commodity are visited, increase the sales volume of commodity.
Proposed algorithm commonly used comprises recommended engine content-based filtration (Content-Based) algorithm commonly used and collaborative filtering (Item-Based, User-based) in the active way of recommendation.Their ubiquity more complicated use threshold higher, have the knotty problems such as data mining and recommended engine use that very strong limitation can not directly apply to internet scale.The present invention is on the basis of further investigation with regard to the Technologies of Recommendation System in E-Commerce principle; Adopt initiatively recommendation and passive type to recommend the way of recommendation that combines; Combining global is fitst water, and most popular Lucene global search technology has been developed the ecommerce recommend method of own uniqueness.Recommended engine algorithm more complicated commonly used uses threshold higher, has the knotty problems such as data mining and recommended engine use that very strong limitation can not directly apply to internet scale.The present invention combines the relative merits of information filtering algorithm and collaborative filtering, solves these problems well in conjunction with the lucene global search technology.
The present invention includes active recommend method and passive type recommend method, subscribe in the business opportunity on Wang Pu backstage, business opportunity express delivery module has adopted active recommendation.The buyer on Wang Pu backstage recommends, and buyer's inquiry, dealing speed are joined and all adopted the passive type recommendation.
As shown in Figure 1, active recommend method comprises the steps:
Step 1 when user capture, selects to subscribe to type (buying or supply) by the user, sets and subscribes to keyword, selects relevant industries;
Step 2, the subscription condition that the user is set stores database into;
The subscription condition is obtained and analyzed to step 3, extracts the field relevant with the subscription condition (title, industry, time), according to the relevant field acquaintance property coupling and the ordering of extracting;
Step 4 is with the user that recommends after handling.
As shown in Figure 2, the passive type recommend method comprises the steps:
Step 1, when user capture, all operations behavior and related data information in the recording user access process; All operations behavior when utilizing platform monitoring system monitoring and recording user visit enterprise to converge net (comprises and browsing; Inquiry, quotation, comment etc.) and the user enterprise is converged the visit situation of net and client retail shop (comprise visiting and relate to module; Access time, information such as the residence time of the page).
Step 2 is put in order user data, forms data warehouse.Monitor data is submitted to recommended engine, classified by recommended engine, data preparations such as polymerization operation forms data warehouse.
Step 3 is carried out mining analysis to the data of different types in the data warehouse, and extracts representative content characteristic values.
Step 4 is formulated recommendation rules, and sets relevant data field (like name of product, product classification, Product labelling, product summary).
Step 5 is called different recommendation rules according to different recommended requirements, according to recommendation rules relevant data content eigenwert is carried out correlation analysis and knowledge and magnanimity matching operation mutually;
Step 6, recommended engine sorts to the data after mating according to associated weight, and the client is given in the data push after will sorting.The buyer recommends regularly from recommended engine, to obtain recommending data with buyer's inquiry module according to different recommended requirements.
Explanation is at last; Above embodiment is only unrestricted in order to technical scheme of the present invention to be described; Other modifications that those of ordinary skills make technical scheme of the present invention perhaps are equal to replacement; Only otherwise break away from the spirit and the scope of technical scheme of the present invention, all should be encompassed in the middle of the claim scope of the present invention.
Claims (3)
1. ecommerce recommend method, it is characterized in that: comprise active recommend method and passive type recommend method, said active recommend method comprises the steps:
Step 1 during user capture, selects to subscribe to type by the user, sets and subscribes to keyword, selects relevant industries;
Step 2, the subscription condition that the user is set stores database into;
Step 3 is extracted the corresponding relevant field of subscription condition, according to relevant field acquaintance property coupling and ordering;
Step 4 is with the user that recommends after handling.
2. a kind of ecommerce recommend method according to claim 1 is characterized in that: said passive type recommend method comprises the steps:
Step 1, when user capture, all operations behavior and related data information in the recording user access process;
Step 2 is put in order user data, forms data warehouse;
Step 3 is analyzed the data in the data warehouse, and extracts representative content characteristic values;
Step 4 is formulated recommendation rules, and sets relevant data field;
Step 5 is carried out correlation analysis and knowledge and magnanimity matching operation mutually according to recommendation rules to relevant data content eigenwert;
Step 6 the weight information of the data after the coupling according to content characteristic values sorted, and the data after will sorting is recommended to show the client.
3. a kind of ecommerce recommend method according to claim 2 is characterized in that: the user data in the said passive type recommend method step 2 comprises supply data, businessman's data and prosperous shop click data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012103485880A CN102831543A (en) | 2012-09-19 | 2012-09-19 | E-commerce recommendation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012103485880A CN102831543A (en) | 2012-09-19 | 2012-09-19 | E-commerce recommendation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102831543A true CN102831543A (en) | 2012-12-19 |
Family
ID=47334662
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012103485880A Pending CN102831543A (en) | 2012-09-19 | 2012-09-19 | E-commerce recommendation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102831543A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104516924A (en) * | 2013-10-08 | 2015-04-15 | 拓广科技股份有限公司 | Demand management method and system |
| CN107122397A (en) * | 2017-03-15 | 2017-09-01 | 百度在线网络技术（北京）有限公司 | Content recommendation method and device |
| CN107329988A (en) * | 2017-06-05 | 2017-11-07 | 国政通科技股份有限公司 | A kind of method and system that data are provided |
| CN109242640A (en) * | 2018-09-27 | 2019-01-18 | 唐山市奎星电商供应链管理有限责任公司 | Electric business supply system and method |
| CN109255682A (en) * | 2018-09-11 | 2019-01-22 | 广东布田电子商务有限公司 | A kind of mixed recommendation system towards electronic business system |
| CN110148032A (en) * | 2019-04-15 | 2019-08-20 | 平安普惠企业管理有限公司 | Products Show method, apparatus, storage medium and server based on geographical location |
| CN111654427A (en) * | 2013-11-22 | 2020-09-11 | 杭州惠道科技有限公司 | Social media system |
| CN112950306A (en) * | 2020-12-31 | 2021-06-11 | 广东华际友天信息科技有限公司 | Recommendation scheme definition method, device, medium and equipment |
| CN114066093A (en) * | 2021-11-25 | 2022-02-18 | 上海数据交易中心有限公司 | Data product processing method for digital business service |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101686249A (en) * | 2008-09-27 | 2010-03-31 | 华为技术有限公司 | Subscription method and system of recommended information and recommended service server |
| KR20120021387A (en) * | 2010-07-29 | 2012-03-09 | 주식회사 넷스루 | System and method for calculating customer preference score based on behavioral idata of web site users |
| CN102663627A (en) * | 2012-04-26 | 2012-09-12 | 焦点科技股份有限公司 | Personalized recommendation method |
-
2012
- 2012-09-19 CN CN2012103485880A patent/CN102831543A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101686249A (en) * | 2008-09-27 | 2010-03-31 | 华为技术有限公司 | Subscription method and system of recommended information and recommended service server |
| KR20120021387A (en) * | 2010-07-29 | 2012-03-09 | 주식회사 넷스루 | System and method for calculating customer preference score based on behavioral idata of web site users |
| CN102663627A (en) * | 2012-04-26 | 2012-09-12 | 焦点科技股份有限公司 | Personalized recommendation method |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104516924A (en) * | 2013-10-08 | 2015-04-15 | 拓广科技股份有限公司 | Demand management method and system |
| CN111654427A (en) * | 2013-11-22 | 2020-09-11 | 杭州惠道科技有限公司 | Social media system |
| CN107122397A (en) * | 2017-03-15 | 2017-09-01 | 百度在线网络技术（北京）有限公司 | Content recommendation method and device |
| CN107329988A (en) * | 2017-06-05 | 2017-11-07 | 国政通科技股份有限公司 | A kind of method and system that data are provided |
| CN109255682A (en) * | 2018-09-11 | 2019-01-22 | 广东布田电子商务有限公司 | A kind of mixed recommendation system towards electronic business system |
| CN109242640A (en) * | 2018-09-27 | 2019-01-18 | 唐山市奎星电商供应链管理有限责任公司 | Electric business supply system and method |
| CN110148032A (en) * | 2019-04-15 | 2019-08-20 | 平安普惠企业管理有限公司 | Products Show method, apparatus, storage medium and server based on geographical location |
| CN112950306A (en) * | 2020-12-31 | 2021-06-11 | 广东华际友天信息科技有限公司 | Recommendation scheme definition method, device, medium and equipment |
| CN114066093A (en) * | 2021-11-25 | 2022-02-18 | 上海数据交易中心有限公司 | Data product processing method for digital business service |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109359244B (en) | Personalized information recommendation method and device | |
| CN102831543A (en) | E-commerce recommendation method | |
| CN109685631B (en) | Personalized recommendation method based on big data user behavior analysis | |
| Wu et al. | Turning clicks into purchases: Revenue optimization for product search in e-commerce | |
| CN105589905B (en) | The analysis of user interest data and collection system and its method | |
| Sivapalan et al. | Recommender systems in e-commerce | |
| US9400831B2 (en) | Providing information recommendations based on determined user groups | |
| CN104035927B (en) | Search method and system based on user behaviors | |
| CN104679771B (en) | A kind of individuation data searching method and device | |
| CN103914478B (en) | Webpage training method and system, webpage Forecasting Methodology and system | |
| CN102591995B (en) | Processing method and device based on user information of cloud data center | |
| CN105183727A (en) | Method and system for recommending book | |
| CN105894332A (en) | Commodity recommendation method, device and system based on user behavior analysis | |
| CN103246980A (en) | Information output method and server | |
| CN102411754A (en) | Personalized recommendation method based on commodity property entropy | |
| CN106708821A (en) | User personalized shopping behavior-based commodity recommendation method | |
| CN101460942A (en) | Method and system for computerized searching and matching using emotional preference | |
| CN109615432A (en) | Consumer behaviour portrait tool based on big data | |
| CN108614832A (en) | A kind of user individual commercial articles searching implementation method and device | |
| CN104462336A (en) | Information pushing method and device | |
| CN104391999A (en) | Information recommendation method and device | |
| Wang et al. | Connecting users with similar interests via tag network inference | |
| CN109299426A (en) | A kind of recommended method and device of accurate top information | |
| CN108446333B (en) | Big data text mining processing system and method thereof | |
| KR101758695B1 (en) | Multi korean wave contents search and recommendation system based on social taste anaysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20121219 |