CN107895303A - A kind of method of the personalized recommendation based on OCEAN models - Google Patents
A kind of method of the personalized recommendation based on OCEAN models Download PDFInfo
- Publication number
- CN107895303A CN107895303A CN201711131237.3A CN201711131237A CN107895303A CN 107895303 A CN107895303 A CN 107895303A CN 201711131237 A CN201711131237 A CN 201711131237A CN 107895303 A CN107895303 A CN 107895303A
- Authority
- CN
- China
- Prior art keywords
- user
- models
- ocean
- personalized recommendation
- targeted customer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of method of the personalized recommendation based on OCEAN models, by establishing the OCEAN models of microblog users, the personalized recommendation method based on user's OCEAN models is realized.When the OCEAN models of user are established, the microblogging text of user is imported into LDA models, finds implicit implied meaning from text with unsupervised approach, improves the precision of prediction.Meanwhile establish personalized recommendation on the basis of user clustering, the hunting zone of user is reduced, reduces the amount of calculation of real-time recommendation.Studied with reference to the OCEAN models of user in personalized recommendation, the character trait for being deep into user, the psychology of user is more conformed to during personalized recommendation, there is the higher degree of accuracy.
Description
Technical field
The invention belongs to personality prediction and personalized recommendation technical field, more specifically, is related to one kind and is based on OCEAN
The method of the personalized recommendation of model.
Background technology
In psychology, OCEAN models are for describing the five of mankind's personality extensive dimensions, and this theory is based on big
Five personality factors models.Five class factors of OCEAN models include：Preciseness, extropism, opening, pleasant property and neurotic people
Lattice speciality.O represents Openness to experience (opening), and C represents Conscientiousness (preciseness), E generations
Table Extraversion (extropism), A represent Agreeableness (pleasant property), and N represents Neuroticism (nervousness).This
Five kinds of factors provide abundant conceptual framework.And the research of forefathers is found, five-factor model personality theoretical model is with people in social activity
Strong association be present in the behavior of website.
Current personalized recommendation algorithm can substantially be divided into four classes：
(1) demographic recommendation mechanisms are based on, are a kind of most readily achieved recommendation methods, it is simple root
The degree of correlation of user is found according to the essential information of system user, other articles for then liking similar users are recommended currently
User.
(2) content-based recommendation, is the recommendation mechanisms that are most widely used at the beginning of recommended engine occurs, its core
Thought is that the correlation of article or content is found according to the metadata for recommending article or content, and it is conventional to be then based on user
Hobby record, recommends the similar article of user.This commending system is used in the application of some information classes, for article sheet
Body extracts keyword of some labels as this article, and the similarity of two articles can be then evaluated by these labels.
(3) recommendation based on correlation rule, it is more often seen in e-commerce system, and is also proved to effective.
Its actual meaning is that the user that have purchased some articles is more likely to buy other articles.Recommendation system based on correlation rule
The primary goal of system is to excavate correlation rule, that is, those article set for being bought by many users simultaneously, these set
Interior article can mutually be recommended.
(4) collaborative filtering, it is a kind of widely used recommendation method in commending system.This algorithm is based on one
" Things of a kind come together " it is assumed that liking the user of identical items more likely has identical interest.Based on cooperateing with
The commending system of filter is generally used among the system of user's scoring, goes to portray hobby of the user for article by fraction.
Collaborative filtering is considered as the model using group wisdom, it is not necessary to carries out specially treated to project, but establishes thing by user
Contacting between product and article.At present, Collaborative Filtering Recommendation System is divided into two types：Based on user (User-based)
Recommendation and recommendation based on article (Item-based).
However, current personalized recommendation method is substantially four classes based on more than, not well with reference to user's
Character trait is marketed.The behavior of user is not random, but contains many specific patterns.The network social intercourse of user
Behavior reflects user's personality, while the personality of user also contributes to user behavior, therefore in online precision marketing, online commodity
The personality of user can be taken into account when recommendation, social recommendation and auxiliary product design, obtain more preferable result.
The content of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide a kind of personalized recommendation based on OCEAN models
Method, the carry out personalized recommendation based on user's personality.
For achieving the above object, a kind of method of the personalized recommendation based on OCEAN models of the present invention, its feature exist
In comprising the following steps：
(1) the OCEAN models of social network sites user, are established
(1.1) some microblogging accounts, are chosen, five kinds of Personality tests are carried out to these users, obtain five kinds of dimensions'
Score, then the OCEAN models using the score of this five kinds of dimensions as subject user；
(1.2) content of pages, is obtained by way of simulation browser, the microblog data of crawl subject user, respectively will
The microblog data of every user is aggregated into a text document；
(1.3), text document is pre-processed：Text document is filtered, word segmentation processing, and is deposited after removing stop-word
It is placed in the database specified；
(1.4), the text document of all subject users in database is imported into LDA topic models, LDA topic models
Export the text document theme probability distribution of every subject user；
(1.5), the document subject matter probability distribution to be tested user inputs as sample, to be tested the OCEAN models of user
Exported as sample, be trained, established between customer documentation theme distribution and user's OCEAN models using BP neural network
Mapping model, and OCEAN model of the mapping model as prediction social network sites user；
(2) personalized recommendation, is carried out to user based on the OCEAN models of social network sites user
(2.1), user clustering
OCEAN models based on social network sites user, user is divided into K kind different characters using K mean cluster algorithm
Customer group；
(2.2) personalized recommendation, is carried out to it according to targeted customer's generic
When targeted customer occurs, it is first determined the cluster classification where targeted customer, then by class where targeted customer
All microbloggings of each user hair in not recycle term frequency-inverse document respectively to each respectively as a Candidate Set item
Text feature is carried out in Candidate Set item to randomly select, constructs a n-dimensional vector, and the attribute as each Candidate Set item provides
Material, wherein, a microblogging is often extracted as one-dimensional vector；
One text document is aggregated into according to the microblog data of targeted customer, also with term frequency-inverse document frequency, to mesh
The text document of mark user carries out text feature and randomly selected, and constructs a m dimensional vector, and the hobby as targeted customer provides
Material；
According to cosine similarity formula, the hobby data and the phase of each Candidate Set item attribute data of user are calculated
Like degree, using similarity highest Candidate Set item as recommending to collect, targeted customer is recommended.
What the goal of the invention of the present invention was realized in：
A kind of method of the personalized recommendation based on OCEAN models of the present invention, by the OCEAN moulds for establishing microblog users
Type, realize the personalized recommendation method based on user's OCEAN models.When the OCEAN models of user are established, by user's
Microblogging text is imported into LDA models, finds implicit implied meaning from text with unsupervised approach, improves the accurate of prediction
Degree.Meanwhile establish personalized recommendation on the basis of user clustering, the hunting zone of user is reduced, reduces real-time recommendation
Amount of calculation.With reference to user OCEAN models in personalized recommendation, the character trait for being deep into user is studied, in personalization
The psychology of user is more conformed to during recommendation, there is the higher degree of accuracy.
Meanwhile a kind of method of the personalized recommendation based on OCEAN models of the present invention also has the advantages that：
(1) the OCEAN models of microblog users, are established, user's personality this is considered before traditional personalized recommendation
Individual index, the personality of user and the hobby of user are combined, not only the degree of accuracy is higher for such recommendation method, and is more bonded
The psychology of user.
(2), when being clustered to user, the selection of the initial cluster center of clustering algorithm is not random, people
Work chooses the higher user of microblogging homepage visit capacity as cluster centre, can preferably reduce isolated point.
Brief description of the drawings
Fig. 1 is a kind of method flow diagram of the personalized recommendation based on OCEAN models of the present invention；
Fig. 2 is LDA topic model figures.
Embodiment
The embodiment of the present invention is described below in conjunction with the accompanying drawings, so as to those skilled in the art preferably
Understand the present invention.Requiring particular attention is that in the following description, when known function and the detailed description of design perhaps
When can desalinate the main contents of the present invention, these descriptions will be ignored herein.
Embodiment
Fig. 1 is a kind of method flow diagram of the personalized recommendation based on OCEAN models of the present invention.
In the present embodiment, as shown in figure 1, a kind of method of the personalized recommendation based on OCEAN models of the present invention, including
Following steps：
S1 chooses some microblogging accounts, carries out five kinds of Personality tests to these users, obtains the score of five kinds of dimensions,
OCEAN models using the score of this five kinds of dimensions as subject user again；
In the present embodiment, using 1991, Univ California-Berkeley psychologist Oliver P John exist
The Big five inventory (Big Five Inventory, BFI) worked out on the basis of OCEAN model theories is to obtain now generally
The Personality test scale of accreditation, the reliability and validity of this scale have all obtained extensive checking in multinomial Experiment of Psychology, this
Application employs this scale, to obtain user's OCEAN models required for training.
S2, content of pages, the microblog data of crawl subject user, the microblogging of user are obtained by way of simulation browser
Data are divided into two parts：Text document and user basic information.Text document refers to collecting for all microblogging texts of user's hair,
User basic information includes the user's registration time, user pays close attention to quantity, user's microblogging bar number, whether has individualized signature etc., then
The microblog data of every user is aggregated into a text document respectively；
S3, text document is pre-processed：Text document is filtered, word segmentation processing, and is deposited after removing stop-word
In specified database；
S4, by database it is all subject user text document imported into LDA topic models, LDA topic models are defeated
Go out the text document theme probability distribution of every subject user；
In the present embodiment, LDA topic models are as shown in Fig. 2 parameter definition is as shown in table 1 in LDA topic models；
Symbolic interpretation：
Table 1
The input of LDA topic models：The set of all user version documents, number of topics K, hyper parameter α and β are according to common
Empirical value：K=10 is set,β=0.01, γ=20
The output of LDA topic models：The theme probability distribution of each user version document.
S5, inputted as sample using being tested the document subject matter probability distribution of user, using be tested the OCEAN models of user as
Sample exports, and is trained using BP neural network, the mapping established between customer documentation theme distribution and user's OCEAN models
Model, and OCEAN model of the mapping model as prediction social network sites user；
S6, the cluster based on social network sites user
OCEAN models based on social network sites user, user is divided into K kind different characters using K mean cluster algorithm
Customer group；
In the present embodiment, k-means clustering algorithms efficiency high, extensively should when being clustered to large-scale data
With, and have good effect on low data collection.The present invention selects k-means clustering algorithms.
If k is the input parameter of k-means algorithms, the number exported after the algorithm is split and calculated on data set is represented
Amount, data set are made up of n data point, represent the quantity of all users, and input parameter is the number k and user of cluster
OCEAN model datas.Specific algorithm is as follows：
1) by several set I={ i of data of five dimensions of user's OCEAN models1,i2,...,i5}；
2) m all users is retrieved, is designated as set U={ u1,u2,...,um}；
3) it is artificial to choose the wherein different user of its higher label of visit capacity as in initial cluster from m user
The heart, it is designated as { W1,W2,...,WK}；
4) input vector is circulated, calculates the average value of object in each cluster, cluster centre is updated, until no longer becoming
Change.
S7, personalized recommendation carried out to it according to targeted customer's generic
When targeted customer occurs, it is first determined the cluster classification where targeted customer, then by class where targeted customer
All microbloggings of each user hair in not recycle term frequency-inverse document respectively to each respectively as a Candidate Set item
Text feature is carried out in Candidate Set item to randomly select, constructs a n-dimensional vector, and the attribute as each Candidate Set item provides
Material, wherein, a microblogging is often extracted as one-dimensional vector；
Such as：Remember that the collection for all microblogging Candidate Sets being collected into is combined into D={ d1,d2,...,dN, occur in all microbloggings
The collection of word be combined into T={ t1,t2,...,tN}.That is, we have a N pieces Candidate Set item to be processed, and these item
In contain the different words of n.We finally will represent that an item, such as jth piece item are expressed using a vector
For dj={ w1j,w2j,...,wnj, wherein w1jRepresent the 1st word t1Weight in article j, it is more important to be worth bigger expression；Institute
With in order to represent jth piece item, it is necessary to calculate djThe value of each component.Utilize the term frequency-inverse document frequency commonly used in information retrieval
(term frequency-inverse document frequency, abbreviation tf-idf).In jth piece microblogging with kth in dictionary
Tf-idf corresponding to individual word is：
Wherein TF (tk,dj) it is the number that k-th of word occurs in Candidate Set item j, and nkIt is that all microbloggings include
The microblogging quantity of k-th of word.
Weight of final k-th of the word in microblogging j is obtained by following formula：
One text document is aggregated into according to the microblog data of targeted customer, also with term frequency-inverse document frequency, to mesh
The text document of mark user carries out text feature and randomly selected, and constructs a m dimensional vector, and the hobby as targeted customer provides
Material；
According to cosine similarity formula, the hobby data and the phase of each Candidate Set item attribute data of user are calculated
Like degree, using similarity highest Candidate Set item as recommending to collect, targeted customer is recommended.
Wherein, cosine similarity formula is：
If scorings of the user U and candidate items I in n dimensions project spatially is expressed as vectorial Ua、Ia, then it is similar
Property cos (U, I) is：
UaHobby values of the targeted customer U to a items is represented, i.e., is worth in hobby data corresponding to a items.IaRepresent that candidate waits
It is worth in selected works item corresponding to a items.
Although the illustrative embodiment of the present invention is described above, in order to the technology of the art
Personnel understand the present invention, it should be apparent that the invention is not restricted to the scope of embodiment, to the common skill of the art
For art personnel, if various change in the spirit and scope of the present invention that appended claim limits and determines, these
Change is it will be apparent that all utilize the innovation and creation of present inventive concept in the row of protection.
Claims (1)
- A kind of 1. method of the personalized recommendation based on OCEAN models, it is characterised in that comprise the following steps：(1) the OCEAN models of social network sites user, are established(1.1) some microblogging accounts, are chosen, five kinds of Personality tests are carried out to these users, obtain the score of five kinds of dimensions, OCEAN models using the score of this five kinds of dimensions as subject user again；(1.2) content of pages, the microblog data of crawl subject user, respectively by every, are obtained by way of simulation browser The microblog data of user is aggregated into a text document；(1.3), text document is pre-processed：Text document is filtered, word segmentation processing, and is stored in after removing stop-word In the database specified；(1.4), the text document of all subject users in database is imported into LDA topic models, the output of LDA topic models The text document theme probability distribution of every subject user；(1.5), inputted using being tested the document subject matter probability distribution of user as sample, using be tested the OCEAN models of user as Sample output sample output, is trained using BP neural network, establish customer documentation theme distribution and user OCEAN models it Between mapping model, and the mapping model as prediction social network sites user OCEAN models；(2) personalized recommendation, is carried out to user based on the OCEAN models of social network sites user(2.1), user clusteringOCEAN models based on social network sites user, user is divided into the user of K kind different characters using K mean cluster algorithm Group；(2.2) personalized recommendation, is carried out to it according to targeted customer's genericWhen targeted customer occurs, it is first determined the cluster classification where targeted customer, then by classification where targeted customer Each user hair all microbloggings respectively as a Candidate Set item, recycle term frequency-inverse document respectively to each candidate Text feature is carried out in collection item to randomly select, and constructs a n-dimensional vector, as each Candidate Set item attribute data, Wherein, a microblogging is often extracted as one-dimensional vector；One text document is aggregated into according to the microblog data of targeted customer, also with term frequency-inverse document frequency, target used The text document at family carries out text feature and randomly selected, and constructs a m dimensional vector, the hobby data as targeted customer；According to cosine similarity formula, the hobby data for calculating user is similar to each Candidate Set item attribute data Degree, using similarity highest Candidate Set item as recommending to collect, recommend targeted customer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711131237.3A CN107895303B (en) | 2017-11-15 | 2017-11-15 | Personalized recommendation method based on OCEAN model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711131237.3A CN107895303B (en) | 2017-11-15 | 2017-11-15 | Personalized recommendation method based on OCEAN model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107895303A true CN107895303A (en) | 2018-04-10 |
| CN107895303B CN107895303B (en) | 2022-03-25 |
Family
ID=61804442
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711131237.3A Active CN107895303B (en) | 2017-11-15 | 2017-11-15 | Personalized recommendation method based on OCEAN model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107895303B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109593844A (en) * | 2019-01-09 | 2019-04-09 | 首慈康健养老有限公司 | A kind of kit and detection method detecting neurotic personality gene |
| CN109628610A (en) * | 2019-01-09 | 2019-04-16 | 首慈康健养老有限公司 | Detect the kit and detection method of neurotic personality gene |
| CN109766493A (en) * | 2018-12-24 | 2019-05-17 | 哈尔滨工程大学 | A kind of cross-domain recommended method combining personality characteristics under neural network |
| CN110457590A (en) * | 2019-06-25 | 2019-11-15 | 华院数据技术(上海)有限公司 | Intelligent subscriber portrait method based on small data input |
| CN110647678A (en) * | 2019-09-02 | 2020-01-03 | 杭州数理大数据技术有限公司 | Recommendation method based on user character label |
| CN111125469A (en) * | 2019-12-09 | 2020-05-08 | 重庆邮电大学 | User clustering method and device for social network and computer equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778260A (en) * | 2014-03-03 | 2014-05-07 | 哈尔滨工业大学 | Individualized microblog information recommending system and method |
| CN105701210A (en) * | 2016-01-13 | 2016-06-22 | 福建师范大学 | Microblog theme emotion analysis method based on mixed characteristic calculation |
| CN105760547A (en) * | 2016-03-16 | 2016-07-13 | 中山大学 | Book recommendation method and system based on user clustering |
-
2017
- 2017-11-15 CN CN201711131237.3A patent/CN107895303B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778260A (en) * | 2014-03-03 | 2014-05-07 | 哈尔滨工业大学 | Individualized microblog information recommending system and method |
| CN105701210A (en) * | 2016-01-13 | 2016-06-22 | 福建师范大学 | Microblog theme emotion analysis method based on mixed characteristic calculation |
| CN105760547A (en) * | 2016-03-16 | 2016-07-13 | 中山大学 | Book recommendation method and system based on user clustering |
Non-Patent Citations (8)
| Title |
|---|
| HUGO HROMIC等: "Graph-Based Methods for Clustering Topics of Interest in Twitter", 《LECTURE NOTES IN COMPUTER SCIENCE》 * |
| MAYURI PUNDLIK KALGHATGI等: "A Neural Network Approach to Personality Prediction based on the Big-Five Model", 《INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN ADVANCED ENGINEERING (IJIRAE)》 * |
| 唐沁钦: "多媒体系统中个性化推荐的研究和设计", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
| 安悦等: "基于内容的热门微话题个性化推荐研究", 《情报杂志》 * |
| 宋志理: "基于LDA模型的文本分类研究", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
| 应晓敏等: "条条大路通罗马—Internet个性化服务的主要形式", 《计算机世界》 * |
| 马如江: "基于Internet的音乐推荐系统的设计与实现", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
| 高斐: "面向海量数据环境的个性化推荐机制应用研究", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766493A (en) * | 2018-12-24 | 2019-05-17 | 哈尔滨工程大学 | A kind of cross-domain recommended method combining personality characteristics under neural network |
| CN109766493B (en) * | 2018-12-24 | 2022-08-02 | 哈尔滨工程大学 | Cross-domain recommendation method combining personality characteristics under neural network |
| CN109593844A (en) * | 2019-01-09 | 2019-04-09 | 首慈康健养老有限公司 | A kind of kit and detection method detecting neurotic personality gene |
| CN109628610A (en) * | 2019-01-09 | 2019-04-16 | 首慈康健养老有限公司 | Detect the kit and detection method of neurotic personality gene |
| CN110457590A (en) * | 2019-06-25 | 2019-11-15 | 华院数据技术(上海)有限公司 | Intelligent subscriber portrait method based on small data input |
| CN110457590B (en) * | 2019-06-25 | 2021-08-27 | 华院计算技术(上海)股份有限公司 | Intelligent user portrait drawing method based on small data input |
| CN110647678A (en) * | 2019-09-02 | 2020-01-03 | 杭州数理大数据技术有限公司 | Recommendation method based on user character label |
| CN110647678B (en) * | 2019-09-02 | 2022-11-15 | 杭州数理大数据技术有限公司 | Recommendation method based on user character label |
| CN111125469A (en) * | 2019-12-09 | 2020-05-08 | 重庆邮电大学 | User clustering method and device for social network and computer equipment |
| CN111125469B (en) * | 2019-12-09 | 2022-06-10 | 重庆邮电大学 | User clustering method and device of social network and computer equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107895303B (en) | 2022-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107895303A (en) | A kind of method of the personalized recommendation based on OCEAN models | |
| Cheng et al. | Personalized click prediction in sponsored search | |
| CN103914478B (en) | Webpage training method and system, webpage Forecasting Methodology and system | |
| CN104199822B (en) | It is a kind of to identify the method and system for searching for corresponding demand classification | |
| CN106339502A (en) | Modeling recommendation method based on user behavior data fragmentation cluster | |
| CN108763362A (en) | Method is recommended to the partial model Weighted Fusion Top-N films of selection based on random anchor point | |
| CN103455487B (en) | The extracting method and device of a kind of search term | |
| CN104199833B (en) | The clustering method and clustering apparatus of a kind of network search words | |
| CN106202294B (en) | Related news computing method and device based on keyword and topic model fusion | |
| CA2632156A1 (en) | Information retrieval system and method using a bayesian algorithm based on probabilistic similarity scores | |
| CN104615779A (en) | Method for personalized recommendation of Web text | |
| Raviv et al. | A ranking framework for entity oriented search using markov random fields | |
| CN108319734A (en) | A kind of product feature structure tree method for auto constructing based on linear combiner | |
| CN107193883B (en) | Data processing method and system | |
| CN110990670B (en) | Growth incentive book recommendation method and recommendation system | |
| CN107423335A (en) | A kind of negative sample system of selection for single class collaborative filtering problem | |
| Islam et al. | Review analysis of ride-sharing applications using machine learning approaches: Bangladesh perspective | |
| Ramesh et al. | Personalized search engine using social networking activity | |
| Xiao | A Survey of Document Clustering Techniques & Comparison of LDA and moVMF | |
| Beheshti-Kashi et al. | Trendfashion-a framework for the identification of fashion trends | |
| Syn et al. | Using latent semantic analysis to identify quality in use (qu) indicators from user reviews | |
| Jiang et al. | Durable product review mining for customer segmentation | |
| Bhojne et al. | Collaborative approach based restaurant recommender system using Naive Bayes | |
| Shuxian et al. | Design and implementation of movie recommendation system based on naive bayes | |
| CN113362034B (en) | Position recommendation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |