CN107862002A - A kind of personalized recommendation system - Google Patents
A kind of personalized recommendation system Download PDFInfo
- Publication number
- CN107862002A CN107862002A CN201710997071.7A CN201710997071A CN107862002A CN 107862002 A CN107862002 A CN 107862002A CN 201710997071 A CN201710997071 A CN 201710997071A CN 107862002 A CN107862002 A CN 107862002A
- Authority
- CN
- China
- Prior art keywords
- user
- usage behavior
- mrow
- behavior data
- information
- 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
- 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
Abstract
The invention discloses a kind of personalized recommendation system, including information acquisition unit, information memory cell, information process unit, information feedback unit and recommend updating block, the system is by introducing forgetting process of the user to the interest of history usage behavior, used each article, which can more accurately be analyzed, is influenceing the importance of the current preference of user, so as to effectively capture the dynamic characteristic of user preference change, and the dynamic change of preference is applied in personalized recommendation, the validity of recommendation results can be lifted.
Description
Technical field
The invention belongs to information recommendation field, more particularly to a kind of personalized recommendation system.
Background technology
With the high speed development of internet, together with daily life is more and more closely connected with internet,
Such as listen music, see a film, do shopping, reading, chatting etc..At the same time, the user of magnanimity and product data are all continuous daily
Ground produces in internet, this cause Internet user be difficult may even not find therefrom rapidly oneself needs or
Unknown message interested.Then, personalized recommendation technology is arisen at the historic moment, and is constantly weeded out the old and bring forth the new.Personalized recommendation technology
It is intended to according to the characteristics of user itself, be modeled the interest preference of user, and and then recommends to meet user individual preference
Information.In order to lift the validity of existing recommendation method, it is necessary to consider the characteristics of preference of user is with time dynamic.Cause
For user history usage behavior will not equably reflect the user given at the time of under preference.
The content of the invention
It is an object of the invention to provide a kind of personalized recommendation system, the system can more accurately catch user preference
Dynamic change, and can be by the interactive feedback mechanism between user, it is current desired to obtain user, so as to by recommendation
As a result it is adjusted correspondingly for user is current desired so that recommendation results can be further bonded needed for user, strengthen user
Experience.
The purpose of the present invention is achieved through the following technical solutions:
A kind of personalized recommendation system, including information acquisition unit, for obtaining the usage behavior data of user;
Information memory cell, the usage behavior data obtained for storage information acquiring unit and history usage behavior number
According to;
Information process unit, for obtaining user profile and Feature Words in usage behavior data, and by the Feature Words
The frequency, time and the region occurred in history usage behavior data is brought into formula (1) and calculates output vector Z, and according to
Output vector Z according to sublist and recommends Feature Words arrangement form sublist according to being recommended and by recommendation results
It is sent to user,
In formula, G1 is characterized the frequency that word occurs in history usage behavior data, and G2 is characterized word and uses row in history
For the time occurred in data, it is to take 10 in nearly three months when the time, takes 6 in nearly 1 year, take 2, G3 to be characterized word the year before and exist
The region occurred in history usage behavior data, often it is 1 in region, is seldom that -1, ω is weighted value in region, ω 1=
{ 0.51,0.17 }, the frequency are more than 100 weighting 0.51, taken less than 100 weighting 0.17, ω 2={ 0.946,0.735,0.308 }
0.946 is weighted when 10,0.735 is weighted when taking 6,0.308, ω 3={ 0.62, -0.24 } are weighted when taking 2,0.62 is weighted when taking 1,
Weighting -0.24 when taking -1, α, β, γ are constant, are 200,10,5 respectively;
Information feedback unit, for receiving interactive feedback operation of the user for the recommendation foundation of recommendation results;
Recommend updating block, user is extracted according to the interactive feedback operation of information feedback unit and clicks on making for recommendation results
With behavioral data, and usage behavior data are sent to information process unit, and form new sublist renewal recommendation results,
Wherein, it is described to recommend according to the Crowds Distribute information for referring at least one dimension corresponding with Feature Words, the people
Group's distributed intelligence is directed to the ratio shared by Feature Words for the crowd of at least one dimension.
Further, the user profile includes user basic information, user interest and User Status.
Further, described information processing unit receives the usage behavior data for recommending updating block to send and extracts use
Feature Words in behavioral data, the Feature Words rank the first in the sub-list.
The invention has the advantages that:
(1) by introducing forgetting process of the user to the interest of historical behavior, it can more accurately analyze and use
Each article influenceing the importance of the current preference of user, it is special so as to effectively capture the dynamic of user preference change
Property, and the dynamic change of preference is applied in personalized recommendation, the validity of recommendation results can be lifted;
(2) can not only automatically obtain personalization recommendation results and its recommend foundation accordingly, and can by with
Interactive feedback mechanism between user, acquisition user is current desired, so as to which the result of recommendation is directed into the current desired progress of user
Corresponding adjustment so that recommendation results are further bonded needed for user, strengthen Consumer's Experience.
Embodiment
The personalized recommendation system that the present embodiment provides includes information acquisition unit, information memory cell, information processing list
Member, information feedback unit and recommendation updating block.
Described information acquiring unit is used for the usage behavior data for obtaining user, and the usage behavior data are believed including user
Breath and Feature Words, user profile include user basic information, user interest and User Status, and Feature Words refer to using in behavioral data
Labeled Feature Words, the user interest that the user interest refers to be come out according to historical behavior data markers are liked, and are specifically
Refer to the semanteme that several Feature Words that Feature Words frequency of occurrence is higher in user's history usage behavior data represent, the user is emerging
Interest can be more more according to the usage behavior data of acquisition more levels off to user and really like.
Described information memory cell is used for the usage behavior data of storage information acquiring unit acquisition and history uses row
For data, data storage is more more more is advantageous to analyze the tendentiousness of user, such as interest, preference.
Described information processing unit, for obtaining user profile and Feature Words in usage behavior data, and by the spy
The frequency, time and the region that sign word occurs in history usage behavior data, which are brought into formula (1), calculates output vector Z, presses
According to output vector Z by Feature Words arrangement form sublist, and according to sublist and foundation is recommended to be recommended and will recommend to tie
Fruit is sent to user,
In formula, G1 is characterized the frequency that word occurs in history usage behavior data, and G2 is characterized word and uses row in history
For the time occurred in data, it is to take 10 in nearly three months when the time, takes 6 in nearly 1 year, take 2, G3 to be characterized word the year before and exist
The region occurred in history usage behavior data, often it is 1 in region, is seldom that -1, ω is weighted value in region, ω 1=
{ 0.51,0.17 }, the frequency are more than 100 weighting 0.51, taken less than 100 weighting 0.17, ω 2={ 0.946,0.735,0.308 }
0.946 is weighted when 10,0.735 is weighted when taking 6,0.308, ω 3={ 0.62, -0.24 } are weighted when taking 2,0.62 is weighted when taking 1,
Weighting -0.24 when taking -1, α, β, γ are constant, are 200,10,5 respectively.
It is described to recommend according to the Crowds Distribute information for referring at least one dimension corresponding with Feature Words, the Crowds Distribute
Information is directed to the ratio shared by Feature Words for the crowd of at least one dimension.
Described information feedback unit is used to receive interactive feedback operation of the user for the recommendation foundation in recommendation results,
The interactive feedback operation can be that in the clicking operation of recommendation results or double-click, dragging etc., other are operated user, only
Want the interactive feedback operation can determine and recommend foundation corresponding to the Feature Words that user chooses.
The recommendation updating block, user is extracted according to the interactive feedback operation of information feedback unit and clicks on recommendation results
Usage behavior data, and usage behavior data are sent to information process unit.Described information processing unit, which receives, to be recommended more
The usage behavior data of new unit transmission simultaneously extract the Feature Words in usage behavior data, and the Feature Words come in the sub-list
First place, do not arranged according still further to the frequency occurred in the historical behavior data, this feature word may be the current preference of user.
The present invention can more accurately be analyzed and used by introducing forgetting process of the user to the interest of historical behavior
Each article crossed is influenceing the importance of the current preference of user, special so as to effectively capture the dynamic of user preference change
Property, and the dynamic change of preference is applied in personalized recommendation, the validity of recommendation results can be lifted.
The present invention can not only automatically obtain the recommendation results of personalization and its recommend foundation accordingly, and can pass through
Interactive feedback mechanism between user, obtain user it is current desired, so as to by the result of recommendation for user it is current desired enter
The corresponding adjustment of row so that recommendation results are further bonded needed for user, strengthen Consumer's Experience.
Described above is only the preferred embodiment of the present invention, but protection scope of the present invention is not limited thereto, any
The transformation and replacement that are carried out based on technical scheme provided by the present invention and inventive concept should all cover the protection model in the present invention
In enclosing.
Claims (3)
- A kind of 1. personalized recommendation system, it is characterised in that including:Information acquisition unit, for obtaining the usage behavior data of user;Information memory cell, the usage behavior data and history usage behavior data obtained for storage information acquiring unit;Information process unit, gone through for obtaining user profile and Feature Words in usage behavior data, and by the Feature Words The frequency, time and the region occurred in history usage behavior data is brought into formula (1) and calculates output vector Z, and according to output Feature Words arrangement form sublist according to sublist and is recommended foundation to be recommended and send recommendation results by vector Z To user,<mrow> <mi>Z</mi> <mo>=</mo> <mrow> <mo>(</mo> <mfrac> <mrow> <mi>&omega;</mi> <mn>1</mn> <mi>G</mi> <mn>1</mn> </mrow> <mi>&alpha;</mi> </mfrac> <mo>+</mo> <mfrac> <mrow> <mi>&omega;</mi> <mn>2</mn> <mi>G</mi> <mn>2</mn> </mrow> <mi>&beta;</mi> </mfrac> <mo>+</mo> <mfrac> <mrow> <mi>&omega;</mi> <mn>3</mn> <mi>G</mi> <mn>3</mn> </mrow> <mi>&gamma;</mi> </mfrac> <mo>)</mo> </mrow> <mo>&times;</mo> <mn>100</mn> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>In formula, G1 is characterized the frequency that word occurs in history usage behavior data, and G2 is characterized word in history usage behavior number According to the time of middle appearance, it is to take 10 in nearly three months when the time, takes 6 in nearly 1 year, take 2, G3 to be characterized word in history the year before The region occurred in usage behavior data, often it is 1 in region, is seldom that -1, ω is weighted value in region, ω 1=0.51, 0.17 }, the frequency is more than 100 weighting 0.51, adds when taking 10 less than 100 weighting 0.17, ω 2={ 0.946,0.735,0.308 } Power 0.946, weights 0.735 when taking 6,0.308, ω 3={ 0.62, -0.24 } is weighted when taking 2,0.62 are weighted when taking 1, when taking -1 Weighting -0.24, α, β, γ are constant, are 200,10,5 respectively;Information feedback unit, for receiving interactive feedback operation of the user for the recommendation foundation of recommendation results;Recommend updating block, the use row of user's click recommendation results is extracted according to the interactive feedback operation of information feedback unit For data, and usage behavior data are sent to information process unit, and form new sublist renewal recommendation results;Wherein, it is described to recommend according to the Crowds Distribute information for referring at least one dimension corresponding with Feature Words, the crowd point Cloth information is directed to the ratio shared by Feature Words for the crowd of at least one dimension.
- 2. personalized recommendation system according to claim 1, it is characterised in that:The user profile is believed substantially including user Breath, user interest and User Status.
- 3. personalized recommendation system according to claim 1, it is characterised in that:Described information processing unit, which receives, to be recommended more The usage behavior data of new unit transmission simultaneously extract the Feature Words in the usage behavior data, and the Feature Words are arranged in the sub-list In first place.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710997071.7A CN107862002A (en) | 2017-10-20 | 2017-10-20 | A kind of personalized recommendation system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710997071.7A CN107862002A (en) | 2017-10-20 | 2017-10-20 | A kind of personalized recommendation system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107862002A true CN107862002A (en) | 2018-03-30 |
Family
ID=61697331
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710997071.7A Pending CN107862002A (en) | 2017-10-20 | 2017-10-20 | A kind of personalized recommendation system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107862002A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108710635A (en) * | 2018-04-08 | 2018-10-26 | 达而观信息科技(上海)有限公司 | A kind of content recommendation method and device |
CN109525480A (en) * | 2018-09-14 | 2019-03-26 | 广东神马搜索科技有限公司 | Customer problem collection system and method |
CN111311295A (en) * | 2018-12-12 | 2020-06-19 | 北京嘀嘀无限科技发展有限公司 | Service mode determining method and device, electronic equipment and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120100867A1 (en) * | 2010-10-25 | 2012-04-26 | Alohar Mobile Inc. | Determining Points of Interest of a Mobile User |
CN103631954A (en) * | 2013-12-13 | 2014-03-12 | 百度在线网络技术(北京)有限公司 | Individualized recommendation method and device |
CN104268155A (en) * | 2014-09-02 | 2015-01-07 | 百度在线网络技术(北京)有限公司 | Immediate mobile terminal personalized recommendation feedback and optimization method and device |
CN104598583A (en) * | 2015-01-14 | 2015-05-06 | 百度在线网络技术(北京)有限公司 | Method and device for generating query sentence recommendation list |
CN105224545A (en) * | 2014-06-03 | 2016-01-06 | 华为技术有限公司 | A kind of position recommend method and device |
CN105677767A (en) * | 2015-12-30 | 2016-06-15 | 联想(北京)有限公司 | Equipment configuration recommending method and device |
-
2017
- 2017-10-20 CN CN201710997071.7A patent/CN107862002A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120100867A1 (en) * | 2010-10-25 | 2012-04-26 | Alohar Mobile Inc. | Determining Points of Interest of a Mobile User |
CN103631954A (en) * | 2013-12-13 | 2014-03-12 | 百度在线网络技术(北京)有限公司 | Individualized recommendation method and device |
CN105224545A (en) * | 2014-06-03 | 2016-01-06 | 华为技术有限公司 | A kind of position recommend method and device |
CN104268155A (en) * | 2014-09-02 | 2015-01-07 | 百度在线网络技术(北京)有限公司 | Immediate mobile terminal personalized recommendation feedback and optimization method and device |
CN104598583A (en) * | 2015-01-14 | 2015-05-06 | 百度在线网络技术(北京)有限公司 | Method and device for generating query sentence recommendation list |
CN105677767A (en) * | 2015-12-30 | 2016-06-15 | 联想(北京)有限公司 | Equipment configuration recommending method and device |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108710635A (en) * | 2018-04-08 | 2018-10-26 | 达而观信息科技(上海)有限公司 | A kind of content recommendation method and device |
CN108710635B (en) * | 2018-04-08 | 2022-03-18 | 达而观信息科技(上海)有限公司 | Content recommendation method and device |
CN109525480A (en) * | 2018-09-14 | 2019-03-26 | 广东神马搜索科技有限公司 | Customer problem collection system and method |
CN109525480B (en) * | 2018-09-14 | 2021-09-10 | 阿里巴巴(中国)有限公司 | User question collection system and method |
CN111311295A (en) * | 2018-12-12 | 2020-06-19 | 北京嘀嘀无限科技发展有限公司 | Service mode determining method and device, electronic equipment and storage medium |
CN111311295B (en) * | 2018-12-12 | 2023-10-13 | 北京嘀嘀无限科技发展有限公司 | Service mode determining method, device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Muneer et al. | Estimation of finite population mean in simple random sampling and stratified random sampling using two auxiliary variables | |
CN107590675A (en) | A kind of recognition methods of user's Shopping Behaviors, storage facilities and mobile terminal based on big data | |
CN102495864A (en) | Collaborative filtering recommending method and system based on grading | |
CN101482884A (en) | Cooperation recommending system based on user predilection grade distribution | |
CN102968506A (en) | Personalized collaborative filtering recommendation method based on extension characteristic vectors | |
CN104077693A (en) | Commodity comparison method, server, client side and e-commerce system | |
CN106897914A (en) | A kind of Method of Commodity Recommendation and system based on topic model | |
CN105824911A (en) | Video recommending method based on LDA user theme model | |
CN107862002A (en) | A kind of personalized recommendation system | |
CN110188268A (en) | A kind of personalized recommendation method based on label and temporal information | |
CN112765480A (en) | Information pushing method and device and computer readable storage medium | |
Lévy-Leduc et al. | Large sample behaviour of some well-known robust estimators under long-range dependence | |
Lee et al. | Recommender system architecture for adaptive green marketing | |
CN103365842B (en) | A kind of page browsing recommends method and device | |
CN109670909A (en) | A kind of travelling products recommended method decomposed based on probability matrix with Fusion Features | |
CN111400613A (en) | Article recommendation method, device, medium and computer equipment | |
CN107341242A (en) | A kind of label recommendation method and system | |
CN114493786A (en) | Information recommendation method and device | |
Shukla et al. | Inventory model for convertible item with deterioration | |
Song et al. | Evaluation model of click rate of electronic commerce advertising based on fuzzy genetic algorithm | |
CN107944946A (en) | Commercial goods labels generation method and device | |
Zhllima et al. | Analysis of consumer preferences for table olives—The case of Albanian urban consumers | |
Cao et al. | Customer demand prediction of service-oriented manufacturing using the least square support vector machine optimized by particle swarm optimization algorithm | |
Borimnejad et al. | Modeling consumer’s behavior for packed vegetable in “Mayadin management organization of Tehran” using artificial neural network | |
CN107944975A (en) | Intention product big data analysis method, apparatus, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180330 |
|
RJ01 | Rejection of invention patent application after publication |