CN106846088A - A kind of Method of Commodity Recommendation of the product electric business website that disappears soon - Google Patents
A kind of Method of Commodity Recommendation of the product electric business website that disappears soon Download PDFInfo
- Publication number
- CN106846088A CN106846088A CN201611198812.7A CN201611198812A CN106846088A CN 106846088 A CN106846088 A CN 106846088A CN 201611198812 A CN201611198812 A CN 201611198812A CN 106846088 A CN106846088 A CN 106846088A
- Authority
- CN
- China
- Prior art keywords
- shops
- consumer
- commodity
- target
- electric business
- 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
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
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/52—Network services specially adapted for the location of the user terminal
Abstract
A kind of Method of Commodity Recommendation of the product electric business website that disappears soon, comprises the following steps：Step one) position data that the registration shops of electric business website platform submits to is stored in LBS cloud storages center；Step 2) then it is target shops with the shops after target consumer marks the commodity in certain shops, the periphery shops for going out target shops using the position data and electronic map software retrieval of LBS cloud storage central storages gathers；Step 3) using periphery set Zhong Ge shops of shops consumer's record information, calculate periphery set Zhong Ge shops of shops consumer and target consumer similarity, it is subdivided go out target consumer similar consumer gather；Step 4) all commodity that the customizable tag in similar consumer's set crosses are extracted, the commodity of the marked mistake of target consumer are therefrom rejected, then calculate interest-degree of the targeted customer to remaining commodity；By interest-degree highest one or more commercial product recommendings to target consumer.
Description
Technical field
The invention belongs to e-commerce field, more particularly to a kind of Method of Commodity Recommendation of the product electric business website that disappears soon.
Background technology
Electric business personalized recommendation is the Characteristic of Interest and buying behavior according to user, to user recommended user letter interested
Breath and commodity.With the continuous expansion of ecommerce scale, commodity number and species rapid growth, customer need spend substantial amounts of
Time can just find the commodity for oneself wanting to buy.This large amount of unrelated information and product process of browsing can undoubtedly make to be submerged in information
Consumer in overload problem is constantly lost in.In order to solve these problems, personalized recommendation system arises at the historic moment.Personalized recommendation
System is built upon a kind of Advanced Business intelligent platform on the basis of mass data is excavated, to help e-commerce website as it is turned round and look at
Visitor's shopping provides completely personalized decision support and information service.
Main proposed algorithm has based on commending contents, collaborative filtering recommending, based on correlation rule is recommended, is pushed away based on effectiveness
Recommend, knowledge based is recommended, combined recommendation.Wherein collaborative filtering recommending technology is using earliest and the most successful in commending system
One of technology.It is general using arest neighbors technology, and the distance between user, Ran Houli are calculated using the history preference information of user
Targeted customer hobby to particular commodity is predicted the weighting evaluation value of commodity evaluation with the nearest-neighbors user of targeted customer
Degree, system according to this fancy grade so as to recommend targeted customer.
Collaborative filtering is based on the assumption that：It is to look for first for a user finds his method of real content interested
To the other users for there are similar interests with this user, then this user is given by their commending contents interested.Its basic thought
It is highly susceptible to understanding, in daily life, we often carry out some selections using the recommendation of good friend.Collaborative filtering is just
It is that this thought is applied to Technologies of Recommendation System in E-Commerce, the evaluation of a certain content is used to target based on other users
Recommended at family.
Commending system based on collaborative filtering can be described as the angle from user to carry out corresponding recommendation, and be automatic
, i.e., the recommendation that user obtains is that system is implicitly obtained from purchasing model or navigation patterns etc., it is not necessary to which user hardy looks for
To the recommendation information for being adapted to oneself interest, some investigation forms are such as filled in.
Traditional Collaborative Filtering Recommendation Algorithm is in the calculating user aggregation process similar to targeted customer's interest, it is necessary to ask
The similarity gone out between targeted customer and every other user, if platform user is huger, this is found and target
The similar user's aggregation process complexity of user interest is O (n*n*d), and n is number of users, and d is the number of commodity.And because
Location-based service is not considered, and the interest difference of the user of different zones may be very big, and this will be considerably increased in collaborative filtering
Sparse Problems, reduce recommendation precision.
The recommendation column of " periphery businessman all orders " of the product electric business platform that disappears soon proposed by the present invention, is the association based on LBS
Same filtering recommendation algorithms, in the searching user aggregation process similar to targeted customer's interest, it is only necessary on targeted customer periphery
User group in find, so the complexity for calculating the user aggregation process similar to targeted customer's interest is substantially reduced, general
Effectively improve the performance of recommended engine.And the hobby of the convenience store on periphery is similar higher, effectively reduces user-project and comment
Sub-matrix it is openness, to a certain extent improve proposed algorithm recommendation precision.
The content of the invention
In order to overcome the deficiencies in the prior art, the invention discloses a kind of commercial product recommending of the product electric business website that disappears soon
Method.
The present invention is achieved through the following technical solutions：
A kind of Method of Commodity Recommendation of the product electric business website that disappears soon, comprises the following steps：
Step one) position data that the registration shops of electric business website platform submits to is stored in LBS cloud storages center；
Step 2) then it is target shops with the shops, using LBS after target consumer marks the commodity in certain shops
The position data and electronic map software retrieval of cloud storage central storage go out the periphery shops set of target shops；
Step 3) using consumer's record information of periphery set Zhong Ge shops of shops, calculate each during periphery shops gathers
The consumer of shops and the similarity of target consumer, it is subdivided go out target consumer similar consumer's set；
Step 4) all commodity that the customizable tag in similar consumer's set crosses are extracted, therefrom reject target consumer
The commodity of the marked mistake of person, then calculate interest-degree of the targeted customer to remaining commodity；By one or more business of interest-degree highest
Product recommend target consumer.
It is further to improve, the step one) in, the position data that the registration shops of electric business website submits to is stored in
LBS cloud storages center, specifically includes following steps：
Registration shops by address be submitted to electric business website platform, the address of electric business website platform examination ＆ verification registration shops letter,
Examination ＆ verification does not pass through, then require to rewrite address, audits again；
After the address examination ＆ verification of registration shops passes through, electric business website platform calls electronic map software, according to the detailed of shops
Address obtains its longitude and latitude；Using the address of shops an as position data, LBS cloud storages center is stored in.
It is further to improve, the step 3) middle disappearing using cosine similarity calculating periphery set Zhong Ge shops of shops
The similarity of the person of expense v and target consumer u, computing formula is：
Wherein u and v is two consumers, and N (u) is the article set of consumer u marks, and N (v) is consumer v marks
Article set.
Further to improve, the step 3) in consumer by similarity more than 0.1893 put similar consumer's collection under
Close.
Further to improve, the step 4) in target consumer u to a calculating for interest-degree p (u, i) of commodity i
Formula is：
WuvIt is the similarity of the consumer v and target consumer u of set Zhong Ge shops of periphery shops；S (u, K) disappears for similar
The person of expense gathers；rviIt is consumer v to the interest of article i, using the hidden feedback data of single behavior, then rvi=1.
It is further to improve, the step 3) in, consumer's record information is consumer's record of nearest month
Information.
Compared with prior art, the present invention has advantages below：
The present invention provides personalized recommendation column for the product electric business platform that disappears soon, in the base of traditional Collaborative Filtering Recommendation Algorithm
LBS is introduced on plinth so that the performance gathered in the searching shops similar to target shops interest-degree is greatly promoted.In view of disappearing soon
The quick circulation of conduct industry, when user-commodity are calculated, only takes the order data of nearest month.According to same geography
The order situation of nearest month on platform of the shops user of commercial circle, is that target shops user recommends shops nearby to enter,
But the commodity set oneself do not ordered.
Brief description of the drawings
Fig. 1 is system construction drawing of the invention；
Fig. 2 is the data interaction flow for registering shops；
Fig. 3 is the flow chart of recommendation of the invention；
Fig. 4 is Similarity Measure schematic diagram of the invention.
Specific embodiment
Embodiment 1
According to the present invention, the recommendation column of a kind of " periphery businessman all orders " of the product electric business platform that disappears soon is designed, be base
In the Collaborative Filtering Recommendation Algorithm of LBS.Fig. 1 illustrates system construction drawing of the invention.The personalization that the present embodiment is based on LBS is pushed away
The system of recommending depends on LBS cloud storages center and user behavior data, recommends the result set for be presented on electric business website UI circle
In the personalized recommendation column of " periphery businessman all orders " in face.The LBS cloud storages center used in the present embodiment is Baidu LBS
Cloud storage center, used electronic map software is Baidu map.
LBS cloud storage central stores the position data that platform registers shops user.So in the product electric business platform note that disappears soon
The shops of volume, requires the better address for filling in shops.The data interaction flow of shops's user's registration is as shown in Figure 2：
1. shops user disappear soon product electric business platform register when, the better address where shops need to be filled in.
2. shops's information of the product electric business that disappears soon platform examination ＆ verification registration, if shops address is filled in not enough in detail, audits not
Pass through, it is desirable to rewrite address.
If 3. the signal auditing of shops passes through, the product electric business platform that disappears soon calls Baidu map api, according to the detailed of shops
Address obtains its longitude and latitude.
4. using shops as a POI (Point Of Information:Position data), it is stored in Baidu's LBS clouds and deposits
Storage center.Baidu LBS. cloud storages are defined as follows shown in table to position data (POI) entities field：
On platform is obtained after the positional information of all registration shops users, it is possible to for target shops user customizes individual character
Change and recommend column " periphery businessman all orders ".The flow of recommendation is as shown in the figure：
(1) periphery shops of searched targets shops set
Because platform all can be stored in Baidu's LBS cloud storages after examination ＆ verification shops information passes through using shops as a POI
Center, so this step can just call the POI Perimeter functions of Baidu map api.
Periphery retrieval refers to centered on a bit (central point is specified by location parameters), to refer near search center point
POI points in set a distance scope (search radius are specified by radius parameters).Retrieval can be specified by tags parameters during retrieval
Type；The sequence (support multilevel sort) of retrieval result is carried out by sortby parameters；It is right that filter parameters can be completed
Specify the screening of data area.
All shops set on 3 kilometers of a certain shops periphery is retrieved such as us, as long as specifying location parameters to be somebody's turn to do
The coordinate of shops, radius parameters are that 3000, tags is that " convenience store " (the tags fields of the POI entities that we create refer to above
It is set to " convenience store "), sortby parameters are appointed as " distance:1 ", i.e., it is ranked up from the near to the remote according to distance.
(2) to find the shops similar to target consumer's interest consumer collected in shops's set on periphery
Finding the shops set similar to target shops consumer interest need to only collect in the shops on target shops periphery now
Found in conjunction.Given consumer u and v, if the article set that N (u) is consumer u likes (or other be marked mode), N
V () is the article set that consumer v likes, then the similarity between cosine similarity can be utilized to calculate u and v.
By taking Fig. 4 as an example, illustrate and calculate consumer's similarity.In this embodiment, consumer A had to article { a, b, d }
Behavior, consumer B had behavior to article { a, c }, and the interest of consumer A and consumer B is calculated using cosine similarity formula
Similarity is：
Similarly we can calculate the similarity of consumer A and consumer C, D：
(3) article is recommended：
If being represented and K target consumer u most like consumer, the thing that consumer in S is liked with set S (u, K)
Product are all extracted, and remove the article that consumer u has liked.For each candidate item i, consumer u is to the emerging of it
Interesting degree is measured with equation below：
rviInterest-degrees of the consumer v to article i is represented, because using the hidden feedback data of single behavior, institute
Some rvi=1.
For example, it is assumed that we will recommend article to consumer A, choose K=3 similar consumer, similar consumer
It is then：B, C, D, then the article that they liked and A did not like has：C, e, then respectively calculate p (A, c) and p (A,
e)：
Finally, the recommendation method is applied to the personalized recommendation column " periphery businessman all orders " of the product electric business website that disappears soon.
Target shops consumer is calculated not ordered within a period of time, but the periphery shops set similar to target shops interest
Have the items list ordered, separate calculate target shops to the level of interest p of each commodity in this items list (u,
I), the height finally according to p (u, i) chooses several commodity and is put into recommendation column.
Above example is merely to illustrate the present invention, but is not limited to the scope of the present invention, every according to of the invention
Any simple modification, equivalent variations and modification that technical spirit is made to following instance, still fall within technical solution of the present invention
In the range of.
Claims (6)
1. a kind of Method of Commodity Recommendation of the product electric business website that disappears soon, it is characterised in that comprise the following steps：
Step one) position data that the registration shops of electric business website platform submits to is stored in LBS cloud storages center；
Step 2) then it is target shops with the shops after target consumer marks the commodity in certain shops, deposited using LBS clouds
The position data and electronic map software retrieval for storing up central storage go out the periphery shops set of target shops；
Step 3) using consumer's record information of periphery set Zhong Ge shops of shops, calculate periphery set Zhong Ge shops of shops
Consumer and target consumer similarity, it is subdivided go out target consumer similar consumer's set；
Step 4) all commodity that the customizable tag in similar consumer's set crosses are extracted, therefrom reject target consumer
Labeled commodity, then calculate interest-degree of the targeted customer to remaining commodity；By interest-degree highest, one or more commodity are pushed away
Recommend to target consumer.
2. a kind of Method of Commodity Recommendation of the product electric business website that disappears soon as claimed in claim 1, it is characterised in that the step
One) in, the position data that the registration shops of electric business website submits to is stored in LBS cloud storages center, is comprised the following steps：
Registration shops by address be submitted to electric business website platform, the address of electric business website platform examination ＆ verification registration shops letter, examination ＆ verification
Do not pass through, then require to rewrite address, audit again；
After the address examination ＆ verification of registration shops passes through, electric business website platform calls electronic map software, according to the better address of shops
Obtain its longitude and latitude；Using the address of shops an as position data, LBS cloud storages center is stored in.
3. a kind of Method of Commodity Recommendation of the product electric business website that disappears soon as claimed in claim 1, it is characterised in that the step
Three) similarity of the consumer v and target consumer u of periphery set Zhong Ge shops of shops, meter are calculated in using cosine similarity
Calculating formula is：
Wherein u and v is two consumers, and N (u) is the article set of consumer u marks, and N (v) is the article of consumer v marks
Set.
4. a kind of Method of Commodity Recommendation of the product electric business website that disappears soon as claimed in claim 1, it is characterised in that the step
Three) consumer in by similarity more than 0.1893 puts similar consumer's set under.
5. a kind of Method of Commodity Recommendation of the product electric business website that disappears soon as claimed in claim 1, it is characterised in that the step
Four) target consumer u is to the computing formula of interest-degree p (u, i) of commodity i in：
WuvIt is the similarity of the consumer v and target consumer u of set Zhong Ge shops of periphery shops；S (u, K) is similar consumer
Set；rviIt is consumer v to the interest of article i, using the hidden feedback data of single behavior, then rvi=1.
6. a kind of Method of Commodity Recommendation of the product electric business website that disappears soon as claimed in claim 1, it is characterised in that the step
Three) in, consumer's record information is consumer's record information of nearest month.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611198812.7A CN106846088A (en) | 2016-12-22 | 2016-12-22 | A kind of Method of Commodity Recommendation of the product electric business website that disappears soon |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611198812.7A CN106846088A (en) | 2016-12-22 | 2016-12-22 | A kind of Method of Commodity Recommendation of the product electric business website that disappears soon |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106846088A true CN106846088A (en) | 2017-06-13 |
Family
ID=59135170
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611198812.7A Pending CN106846088A (en) | 2016-12-22 | 2016-12-22 | A kind of Method of Commodity Recommendation of the product electric business website that disappears soon |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106846088A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110084683A (en) * | 2019-04-29 | 2019-08-02 | 河北商之翼互联网科技有限公司 | A kind of Intelligent Business processing system |
| CN110348921A (en) * | 2018-04-02 | 2019-10-18 | 北京京东尚科信息技术有限公司 | The method and apparatus that shops's article is chosen |
| CN110471936A (en) * | 2019-08-19 | 2019-11-19 | 福建工程学院 | A kind of hybrid SQL automatic scoring method |
| CN111582975A (en) * | 2020-04-23 | 2020-08-25 | 许立达 | Artificial intelligence recommendation method and system based on combination of users, products and advertisements |
| CN112446709A (en) * | 2020-12-15 | 2021-03-05 | 安徽银通物联有限公司 | Face recognition payment authentication method and system |
| CN112513898A (en) * | 2018-07-31 | 2021-03-16 | 株式会社彩 | Alcoholic drink information management system and management method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408647A (en) * | 2014-11-29 | 2015-03-11 | 深圳市无微不至数字技术有限公司 | Credit exchange method, credit exchange device and system |
| CN104462263A (en) * | 2014-11-21 | 2015-03-25 | 厦门雅迅网络股份有限公司 | Method for searching for stores by means of database indexes |
| CN104463637A (en) * | 2014-12-23 | 2015-03-25 | 北京石油化工学院 | Commodity recommendation method and device based on electronic business platform and server |
| CN104504055A (en) * | 2014-12-19 | 2015-04-08 | 常州飞寻视讯信息科技有限公司 | Commodity similarity calculation method and commodity recommending system based on image similarity |
-
2016
- 2016-12-22 CN CN201611198812.7A patent/CN106846088A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104462263A (en) * | 2014-11-21 | 2015-03-25 | 厦门雅迅网络股份有限公司 | Method for searching for stores by means of database indexes |
| CN104408647A (en) * | 2014-11-29 | 2015-03-11 | 深圳市无微不至数字技术有限公司 | Credit exchange method, credit exchange device and system |
| CN104504055A (en) * | 2014-12-19 | 2015-04-08 | 常州飞寻视讯信息科技有限公司 | Commodity similarity calculation method and commodity recommending system based on image similarity |
| CN104463637A (en) * | 2014-12-23 | 2015-03-25 | 北京石油化工学院 | Commodity recommendation method and device based on electronic business platform and server |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110348921A (en) * | 2018-04-02 | 2019-10-18 | 北京京东尚科信息技术有限公司 | The method and apparatus that shops's article is chosen |
| CN110348921B (en) * | 2018-04-02 | 2023-06-02 | 北京京东尚科信息技术有限公司 | Method and device for selecting store articles |
| CN112513898A (en) * | 2018-07-31 | 2021-03-16 | 株式会社彩 | Alcoholic drink information management system and management method |
| CN110084683A (en) * | 2019-04-29 | 2019-08-02 | 河北商之翼互联网科技有限公司 | A kind of Intelligent Business processing system |
| CN110084683B (en) * | 2019-04-29 | 2022-02-11 | 河北商之翼互联网科技有限公司 | Intelligent commercial processing system |
| CN110471936A (en) * | 2019-08-19 | 2019-11-19 | 福建工程学院 | A kind of hybrid SQL automatic scoring method |
| CN110471936B (en) * | 2019-08-19 | 2022-06-07 | 福建工程学院 | Hybrid SQL automatic scoring method |
| CN111582975A (en) * | 2020-04-23 | 2020-08-25 | 许立达 | Artificial intelligence recommendation method and system based on combination of users, products and advertisements |
| CN111582975B (en) * | 2020-04-23 | 2023-06-02 | 许立达 | Artificial intelligence recommendation method and system based on combination of user, product and advertisement |
| CN112446709A (en) * | 2020-12-15 | 2021-03-05 | 安徽银通物联有限公司 | Face recognition payment authentication method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106846088A (en) | A kind of Method of Commodity Recommendation of the product electric business website that disappears soon | |
| US11748379B1 (en) | Systems and methods for generating and implementing knowledge graphs for knowledge representation and analysis | |
| CN108121737B (en) | Method, device and system for generating business object attribute identifier | |
| CN104123332B (en) | The display methods and device of search result | |
| KR100961783B1 (en) | Apparatus and method for presenting personalized goods and vendors based on artificial intelligence, and recording medium thereof | |
| WO2019056661A1 (en) | Search term pushing method and device, and terminal | |
| TW201727558A (en) | System, method, and device for generating a geographic area heat map | |
| CN108701313A (en) | The system and method recommended are generated using data corpus | |
| CN103886001A (en) | Personalized commodity recommendation system | |
| CN102254043A (en) | Semantic mapping-based clothing image retrieving method | |
| CN106960044B (en) | Time perception personalized POI recommendation method based on tensor decomposition and weighted HITS | |
| CN105809475A (en) | Commodity recommendation method compatible with O2O applications in internet plus tourism environment | |
| CN107230098A (en) | Method and system is recommended in a kind of timesharing of business object | |
| CN109460519B (en) | Browsing object recommendation method and device, storage medium and server | |
| US10169797B2 (en) | Identification of entities based on deviations in value | |
| CN104615631B (en) | A kind of method and device of information recommendation | |
| WO2019218654A1 (en) | Product ordering method | |
| CN108563690A (en) | A kind of collaborative filtering recommending method based on object-oriented cluster | |
| KR20170118297A (en) | Method for recommending product based on weather information | |
| CN108733834A (en) | The user oriented recommendation method, apparatus of one kind and storage medium | |
| CN115496566B (en) | Regional specialty recommendation method and system based on big data | |
| CN106708871A (en) | Method and device for identifying social service characteristics user | |
| CN108292409A (en) | Consumer's decision tree generation system | |
| TW201734909A (en) | Method and apparatus for identifying target user | |
| CN110990717B (en) | Interest point recommendation method based on cross-domain association |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170613 |