CN109086281A - A kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm - Google Patents
A kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm Download PDFInfo
- Publication number
- CN109086281A CN109086281A CN201710449140.0A CN201710449140A CN109086281A CN 109086281 A CN109086281 A CN 109086281A CN 201710449140 A CN201710449140 A CN 201710449140A CN 109086281 A CN109086281 A CN 109086281A
- Authority
- CN
- China
- Prior art keywords
- project
- supplier
- scoring
- similarity
- sim
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
Abstract
The data processing system of the purpose of the invention belongs to be specially adapted for the management of specific area, supervision or prediction or the technical field of method, and in particular to a kind of manufacturing industry supplier recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm.The method, in the case where score data is extremely sparse, predicts supplier u in the scoring r ' of project p by step S2U, p, according to this based on acquire similarity, avoid the case where Sparse bring accuracy rate sharply declines.The method of the invention has obtained more accurate supplier's recommendation results according to effective user's similitude;And this algorithm both can be used as manufacturing industry supplier recommendation, can also expand to very much other recommendation fields in the extremely sparse situation of data, such as retail product is recommended, and has wider application range.
Description
Technical field
The data processing system of the purpose of the invention belongs to be specially adapted for the management of specific area, supervision or prediction or side
A kind of technical field of method, and in particular to manufacturing industry supplier recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm.
Background technique
Manufacturing industry faces the huge data of magnanimity, and how preferably analysis mining is that manufacturing industry faces using these information
Ultimate challenge, it is more comprehensive for the former analysis and research at present in terms of client and supplier, but to the analysis of the latter compared with
It is few.With the development of the transition of manufacturing mode and integrated supply chain management theory, as most optimum distribution of resources in manufacturing process
The supplier selection and evaluation problem of classical problem, is constantly endowed new intension, while also having obtained academia and business circles
More and more concerns.
In order to generate accurate recommendation, domestic and international researcher proposes a variety of different proposed algorithms, as collaborative filtering pushes away
Recommend technology, Bayesian network technology, clustering technique, Association Rules Technology etc.;Wherein arest neighbors collaborative filtering recommending is current
Most successful recommended technology, basic thought are that the score data based on the similar nearest-neighbors that score is pushed away to target user's generation
It recommends, since nearest-neighbors are closely similar to the scoring of project and target user, scoring of the target user to non-scoring item
It can be approached by the weighted average that nearest-neighbors score to the project.
But as the expansion of manufacturing industry scale, supplier (user) and project/product data sharply increase, lead to user
The extreme sparsity of score data, in the case where user's score data is extremely sparse, traditional method for measuring similarity exists
Respective drawback, so that the nearest-neighbors inaccuracy for the target user being calculated, the recommendation quality of recommender system sharply decline.
Method for measuring similarity representative at present is that Person associated similarity is related to modified cosine similarity Person
Similitude and modified cosine similarity.On the one hand, such method for measuring similarity is similar between the project of calculating or user
It only considered the data to score jointly when property, cause the user for only possessing common scoring item to have similar possibility, with reality
Border situation is not inconsistent;On the other hand, Collaborative Recommendation is faced with Sparse and is cold-started the challenge of problem, with number of users and item
Mesh number purpose sharply increases, and the calculating accuracy for causing user's nearest-neighbors and item nearest neighbor to occupy reduces, so that recommender system
Recommendation quality sharply decline.In addition, generalling use k Neighborhood Model to the calculating of similarity at present, but needed using k Neighborhood Model
Precompute k nearest-neighbors of user or project, k value will lead to calculation amount when excessive excessive and influence to recommend to generate
Real-time, and k value is too small, will lead to recommendation accuracy decline;Therefore the value of k chooses that how many need will be according to calculating repeatedly
Out, calculating cost has been aggravated.
Therefore traditional method for measuring similarity can not effectively be spent in the case where user's score data is extremely sparse
The similitude between user is measured, so that the nearest-neighbors inaccuracy of the target user calculated, causes entirely to supply
The recommendation quality of the recommended method of quotient sharply declines.
Summary of the invention
The purpose of the present invention is to the deficiencies in the prior art, provide a kind of based on arest neighbors Collaborative Filtering Recommendation Algorithm
Manufacturing industry supplier recommended method, can be in the case where user's score data be extremely sparse effectively between measure user
Similitude, and then improve the recommendation quality to manufacturing industry supplier.
To achieve the above object, the present invention is based on the technologies of supplier's recommended method of arest neighbors Collaborative Filtering Recommendation Algorithm
Scheme is as follows: setting to the project set of supplier u scoring as Iu, the project set to supplier v scoring is Iv, the method
Specific steps are as follows:
S1, it calculates to the union I ' of supplier u and the v project set to scoreuv, i.e. I 'uv=Iu∪Iv;Then supplier u exists
Project space I 'uvIn non-scoring item set NuFor Nu=I 'uv-Iu;
S2, in project space I 'uvIn, to any project p ∈ Nu, scoring R ' of the prediction supplier u in project pu,p, specifically
Step are as follows:
S21, in project space I 'uvIn, the similarity between project i, j is calculated based on Person associated similarity:
Wherein, UiThere is the supply quotient set of scoring for project i;UjThere is the supply quotient set of scoring for project j;Then Uij=Ui
∩Uj, there is the supply quotient set of scoring for project i, j;Ru,iFor supplier u project i scoring;Ru,jIt is supplier u in item
The scoring of mesh j;WithRespectively indicate the average value that all suppliers score in project i and project j;
S22, using with the highest h project of project p similarity as neighbours' project set of project p, i.e., in entire project
Project set Mp={ I is searched in space1,I2,…Im,…,Ih, I1With similitude sim (p, the I of project p1) highest, project I2
With similarity sim (p, the I of project p2) take second place, and so on;
The scoring R ' of S23, prediction supplier u on project pu,p:
Wherein,For project space I 'uvIn all suppliers project p grade average;M indicates project set Mp
Any one of mesh;Sim (p, m) is the similarity of project p and project m;It is supplier n in project space I 'uvIn all items
The average value of mesh scoring;Rn,pFor supplier n project p scoring;
The scoring of S3, the supplier obtained based on S2 in project, using Person associated similarity, calculate supplier u,
Similarity between v:
Wherein, Ru,iFor supplier u project i scoring;Rv,iFor supplier v project i scoring;WithRespectively
Indicate the grade average of supplier u and supplier v on all items;
S4, similarity support weight SW is introduceduv:
Wherein, similarity support SS is the support sample size of scoring similarity, SSuvIndicate phase between supplier u and v
Like degree support, value is the number of entry that supplier's u and v co-production is crossed;SSminAnd SSmaxIt is illustrated respectively in supplier's sky
Between middle similarity support minimum value and maximum value;
S5, to the similarity between supplier u, v multiplied by similarity support weight:
Sim ' (u, v)=sim (u, v) * SWuvFormula 5
S6, scoring R of the supplier u on project i is calculatedu,i, recommendation is generated based on the scoring between supplier:
Wherein, Nk is and the maximum preceding k supplier of supplier u similarity.
The present invention, in the case where score data is extremely sparse, predicts supplier u in the scoring of project p by step S2
R’u,p, according to this based on acquire similarity, avoid the case where Sparse bring accuracy rate sharply declines.
Specifically, heretofore described project is product.The i.e. described project can be supplier's product produced, supply
Scoring of the quotient in project corresponds to the qualification rate that supplier produces certain product.
The beneficial effects of the invention are as follows provide a kind of recommendation side of supplier based on arest neighbors Collaborative Filtering Recommendation Algorithm
Method, this algorithm can effectively measure the similitude between supplier in the case where user's score data is extremely sparse;In turn
More accurate supplier's recommendation results have been obtained according to effective user's similarity;And this algorithm both can be used as manufacturing industry confession
It answers quotient to recommend, other recommendation fields in the extremely sparse situation of data can also be expanded to very much, such as retail product is recommended, tool
There is wider application range.
Specific embodiment
Below with reference to embodiment, implementation of the invention is further described.
In the present invention, if being I to the project set of supplier u scoringu, the project set to supplier v scoring is Iv, institute
State the specific steps of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm are as follows:
S1, it calculates to the union I ' of supplier u and the v project set to scoreuv, i.e. I 'uv=Iu∪Iv;Then supplier u exists
Project space I 'uvIn non-scoring item set NuFor Nu=I 'uv-Iu;
S2, in project space I 'uvIn, to any project p ∈ Nu, scoring R ' of the prediction supplier u in project pu,p, specifically
Step are as follows:
S21, in project space I 'uvIn, the similarity between project i, j is calculated based on Person associated similarity:
Wherein, UiThere is the supply quotient set of scoring for project i;UjThere is the supply quotient set of scoring for project j;Then Uij=Ui
∩Uj, there is the supply quotient set of scoring for project i, j;Ru,iFor supplier u project i scoring;Ru,jIt is supplier u in item
The scoring of mesh j;WithRespectively indicate the average value that all suppliers score in project i and project j;
S22, using with the highest h project of project p similarity as neighbours' project set of project p, i.e., in entire project
Project set Mp={ I is searched in space1,I2,…Im,…,Ih, I1With similitude sim (p, the I of project p1) highest, project I2
With similarity sim (p, the I of project p2) take second place, and so on;
The scoring R ' of S23, prediction supplier u on project pu,p:
Wherein,For project space I 'uvIn all suppliers project p grade average;M indicates project set Mp
Any one of mesh;Sim (p, m) is the similarity of project p and project m;It is supplier n in project space I 'uvIn all items
The average value of mesh scoring;Rn,pFor supplier n project p scoring;
The scoring of S3, the supplier obtained based on S2 in project, using Person associated similarity, calculate supplier u,
Similarity between v:
Wherein, Ru,iFor supplier u project i scoring;Rv,iFor supplier v project i scoring;WithRespectively
Indicate the grade average of supplier u and supplier v on all items;
S4, similarity support weight SW is introduceduv:
Wherein, similarity support SS is the support sample size of scoring similarity, SSuvIndicate phase between supplier u and v
Like degree support, value is the number of entry that supplier's u and v co-production is crossed;SSminAnd SSmaxIt is illustrated respectively in supplier's sky
Between middle similarity support minimum value and maximum value;
S5, to the similarity between supplier u, v multiplied by similarity support weight:
Sim ' (u, v)=sim (u, v) * SWuvFormula 5
S6, scoring R of the supplier u on project i is calculatedu,i, recommendation is generated based on the scoring between supplier:
Wherein, Nk is and the maximum preceding k supplier of supplier u similarity.
The present invention, in the case where score data is extremely sparse, predicts supplier u in the scoring of project p by step S2
r’u,p, according to this based on acquire similarity, avoid the case where Sparse bring accuracy rate sharply declines.
Heretofore described project is product, i.e., the described project can be supplier's product produced, and supplier is in item
Scoring on mesh corresponds to the qualification rate that supplier produces certain product.
Embodiment
It is assumed that there is 8 suppliers to 10 production qualification rate such as the following table 1, User indicates that supplier, Item indicate to produce
Product, number is the qualification rate that supplier produces certain product in table ,/indicate that this supplier does not produce such product temporarily.
18 suppliers of table are to 10 production qualification rates
By taking product I tem6 as an example, only supplier User1 and User7 were produced, and qualification rate is respectively 0.1 and 0.4,
Now more good supplier is being found again, steps are as follows:
S1, it calculates to the union I ' of supplier u and the v project set to scoreuv, that is, calculate supplier has qualification between any two
The union of the project set of rate;As between supplier User1 and User2 project space be (Item1, Item2, Item3,
Item6, Item7, Item8, Item10), between supplier User1 and User3 project space be (Item1, Item2, Item3,
Item4,Item5,Item6,Item7,Item8,Item9);And so on;
S2, in project space I 'uvIn, to any project p ∈ Nu, i.e., the part of any no score data, prediction supplies respectively
Answer scoring R ' of the quotient in the projectu,p, specific steps are as follows:
S21, in project space I 'uvIn, it is similar to sundry item that project Item6 is calculated based on Person associated similarity
Degree, is calculated by formula 1:
Sim (Item6, Item1)=1.0;Sim (Item6, Item2)=0.48564288;sim(Item6,Item3)
=-0.4856429;Sim (Item6, Item4)=0.0;Sim (Item6, Item5)=0.0;Sim (Item6, Item7)=
1.0;Sim (Item6, Item8)=0.0;Sim (Item6, Item9)=1.0;And so on;
S22, using with highest 4 projects of project Item6 similarity as neighbours' project set of project Item6, i.e. Mp
={ Item1, Item7, Item9, Item2 }, I1With similitude sim (p, the I of project p1) highest, project I2It is similar to project p's
Spend sim (p, I2) take second place, and so on;
S23, each and concentrate, predicted respectively using formula 2 supplier User1 project Item4, Item5, Item8,
Scoring on Item9, Item10;Scoring of the supplier User2 on project Item1, Item4, Item5, Item6, Item9,
And so on;
S3, the scoring obtained based on S2 calculate the similarity between supplier using Person associated similarity;It is i.e. sharp
Sim (User1, User2)=0.185135 is acquired with formula 3;Sim (User1, User3)=0.7439289;sim(User1,
User4)=0.81549037, and so on;
S4, similarity support weight SW is introduceduv, i.e., the similarity support between supplier is calculated according to formula 4 and weighed
Weight: SWUser1,User2=0.2857143;SWUser1,User3=0.42857143;SWUser1,User4=0.2857143;And so on;
S5, similarity between supplier is obtained considering after similarity support weight according to formula 5:
Sim (User1, User2)=0.052895717;Sim (User1, User3)=0.31882668;sim(User1,
User4)=0.23299725;And so on
S6, calculated according to formula 6 each supplier production product I tem6 qualification rate successively are as follows: PUser1,Item6=
0.1;PUser2,Item6=0.42;PUser3,Item6=0.34;PUser4,Item6=0.36;PUser5,Item6=0.58;PUser6,Item6=
0.32;PUser7,Item6=0.4;PUser8,Item6=0.64;And then recommendation is generated based on the scoring between supplier: to product I tem6
Consequently recommended more good supplier User8.
It combines above and the present invention is exemplarily described, it is clear that the present invention implements not by the limit of aforesaid way
System, as long as using the improvement for the various unsubstantialities that the inventive concept and technical scheme of the present invention carry out, or not improved general
Conception and technical scheme of the invention directly apply to other occasions, within the scope of the present invention.
Claims (2)
1. a kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm, if to the Item Sets of supplier u scoring
It is combined into Iu, the project set to supplier v scoring is Iv, it is characterised in that:
The specific steps of the method are as follows:
S1, it calculates to the union I ' of supplier u and the v project set to scoreuv, i.e. I 'uv=Iu∪Iv;
Then supplier u is in project space I 'uvIn non-scoring item set NuFor Nu=I 'uv-Iu;
S2, in project space I 'uvIn, to any project p ∈ Nu, scoring R ' of the prediction supplier u in project pu,p, specific steps
Are as follows:
S21, in project space I 'uvIn, the similarity between project i, j is calculated based on Person associated similarity:
Wherein, UiThere is the supply quotient set of scoring for project i;UjThere is the supply quotient set of scoring for project j;Then Uij=Ui∩Uj,
There is the supply quotient set of scoring for project i, j;Ru,iFor supplier u project i scoring;Ru,jIt is supplier u project j's
Scoring;WithRespectively indicate the average value that all suppliers score in project i and project j;
S22, using with the highest h project of project p similarity as neighbours' project set of project p, i.e., in entire project space
Middle lookup project set Mp={ I1,I2,…Im,…,Ih, I1With similitude sim (p, the I of project p1) highest, project I2With item
Similarity sim (p, the I of mesh p2) take second place, and so on;
The scoring R ' of S23, prediction supplier u on project pu,p:
Wherein,For project space I 'uvIn all suppliers project p grade average;M is indicated in project set Mp
Any project;Sim (p, m) is the similarity of project p and project n;It is supplier n in project space I 'uvMiddle all items are commented
The average value divided;Rn,pFor supplier n project p scoring;
The scoring of S3, the supplier obtained based on S2 in project, using Person associated similarity, calculate supplier u, v it
Between similarity:
Wherein, Ru,iFor supplier u project i scoring;Rv,iFor supplier v project i scoring;WithIt respectively indicates
The grade average of supplier u and supplier v on all items;
S4, similarity support weight SW is introduceduv:
Wherein, similarity support SS is the support sample size of scoring similarity, SSuvIndicate similarity between supplier u and v
Support, value are the number of entry that supplier's u and v co-production is crossed;SSminAnd SSmaxIt is illustrated respectively in the supply quotient space
The minimum value and maximum value of similarity support;
S5, to the similarity between supplier u, v multiplied by similarity support weight:
Sim ' (u, v)=sim (u, v) * SWuvFormula 5
S6, scoring R of the supplier u on project i is calculatedu,i, recommendation is generated based on the scoring between supplier:
Wherein, Nk is and the maximum preceding k supplier of supplier u similarity.
2. supplier's recommended method according to claim 1 based on arest neighbors Collaborative Filtering Recommendation Algorithm, feature exist
In: the project is product.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710449140.0A CN109086281A (en) | 2017-06-14 | 2017-06-14 | A kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710449140.0A CN109086281A (en) | 2017-06-14 | 2017-06-14 | A kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109086281A true CN109086281A (en) | 2018-12-25 |
Family
ID=64838933
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710449140.0A Pending CN109086281A (en) | 2017-06-14 | 2017-06-14 | A kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109086281A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110096640A (en) * | 2019-03-08 | 2019-08-06 | 江苏物联网研究发展中心 | User's similarity calculating method in Collaborative Filtering Recommendation System based on classification of the items |
CN112465564A (en) * | 2020-12-11 | 2021-03-09 | 广东电网有限责任公司 | Supplier recommendation method, device and terminal |
CN112488786A (en) * | 2020-11-30 | 2021-03-12 | 广东电网有限责任公司 | Supplier recommendation method and device based on user collaborative filtering |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101329683A (en) * | 2008-07-25 | 2008-12-24 | 华为技术有限公司 | Recommendation system and method |
EP2518679A1 (en) * | 2011-04-26 | 2012-10-31 | YooChoose GmbH | Method and system fo recommending geo-tagged items |
CN102779131A (en) * | 2011-05-12 | 2012-11-14 | 同济大学 | Collaborative filtering recommending method based on multiple-similarity of users |
CN103092911A (en) * | 2012-11-20 | 2013-05-08 | 北京航空航天大学 | K-neighbor-based collaborative filtering recommendation system for combining social label similarity |
CN104933156A (en) * | 2015-06-25 | 2015-09-23 | 西安理工大学 | Collaborative filtering method based on shared neighbor clustering |
CN106326390A (en) * | 2016-08-17 | 2017-01-11 | 成都德迈安科技有限公司 | Recommendation method based on collaborative filtering |
-
2017
- 2017-06-14 CN CN201710449140.0A patent/CN109086281A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101329683A (en) * | 2008-07-25 | 2008-12-24 | 华为技术有限公司 | Recommendation system and method |
EP2518679A1 (en) * | 2011-04-26 | 2012-10-31 | YooChoose GmbH | Method and system fo recommending geo-tagged items |
CN102779131A (en) * | 2011-05-12 | 2012-11-14 | 同济大学 | Collaborative filtering recommending method based on multiple-similarity of users |
CN103092911A (en) * | 2012-11-20 | 2013-05-08 | 北京航空航天大学 | K-neighbor-based collaborative filtering recommendation system for combining social label similarity |
CN104933156A (en) * | 2015-06-25 | 2015-09-23 | 西安理工大学 | Collaborative filtering method based on shared neighbor clustering |
CN106326390A (en) * | 2016-08-17 | 2017-01-11 | 成都德迈安科技有限公司 | Recommendation method based on collaborative filtering |
Non-Patent Citations (2)
Title |
---|
罗辛 等: "通过相似度支持度优化基于K近邻的协同过滤算法", 《计算机学报》 * |
陶维安 等: "基于评分支持度的最近邻协同过滤推荐算法", 《计算机应用研究》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110096640A (en) * | 2019-03-08 | 2019-08-06 | 江苏物联网研究发展中心 | User's similarity calculating method in Collaborative Filtering Recommendation System based on classification of the items |
CN112488786A (en) * | 2020-11-30 | 2021-03-12 | 广东电网有限责任公司 | Supplier recommendation method and device based on user collaborative filtering |
CN112465564A (en) * | 2020-12-11 | 2021-03-09 | 广东电网有限责任公司 | Supplier recommendation method, device and terminal |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Li et al. | Using multidimensional clustering based collaborative filtering approach improving recommendation diversity | |
Lian et al. | A modified colonial competitive algorithm for the mixed-model U-line balancing and sequencing problem | |
CN104317945A (en) | E-commerce website commodity recommending method on basis of search behaviors | |
CN106777051A (en) | A kind of many feedback collaborative filtering recommending methods based on user's group | |
CN106157156A (en) | A kind of cooperation recommending system based on communities of users | |
Li et al. | Improving personalized services in mobile commerce by a novel multicriteria rating approach | |
US10970296B2 (en) | System and method for data mining and similarity estimation | |
Aharon et al. | Off-set: one-pass factorization of feature sets for online recommendation in persistent cold start settings | |
Rokach et al. | Initial profile generation in recommender systems using pairwise comparison | |
CN109086281A (en) | A kind of supplier's recommended method based on arest neighbors Collaborative Filtering Recommendation Algorithm | |
Xu et al. | On the newsvendor model with conditional Value-at-Risk of opportunity loss | |
CN104778237A (en) | Individual recommending method and system based on key users | |
US20200226493A1 (en) | Apparatus and Method for Training a Similarity Model Used to Predict Similarity Between Items | |
Ghuli et al. | A collaborative filtering recommendation engine in a distributed environment | |
Dey et al. | A hybrid fuzzy technique for the selection of warehouse location in a supply chain under a utopian environment | |
CN110209946A (en) | Based on social and community Products Show method, system and storage medium | |
Gupta et al. | A framework for a recommendation system based on collaborative filtering and demographics | |
CN112488863A (en) | Dangerous seed recommendation method and related equipment in user cold start scene | |
CN104063555B (en) | The user model modeling method intelligently distributed towards remote sensing information | |
CN108694234A (en) | A kind of service recommendation model based on improvement collaborative filtering | |
CN106919647B (en) | Clustering-based network structure similarity recommendation method | |
Yan et al. | Optimal inventory policies under stochastic production and demand rates | |
CN109948055A (en) | A kind of recommended method based on group's preference | |
Wang et al. | Using incremental clustering technique in collaborative filtering data update | |
KR102495011B1 (en) | Method for recommending product using regression analysis according to product code and purchase pattern and apparatus for the same |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181225 |