CN108846062B - Method for pushing based on users ' individualized requirement - Google Patents
Method for pushing based on users ' individualized requirement Download PDFInfo
- Publication number
- CN108846062B CN108846062B CN201810562061.5A CN201810562061A CN108846062B CN 108846062 B CN108846062 B CN 108846062B CN 201810562061 A CN201810562061 A CN 201810562061A CN 108846062 B CN108846062 B CN 108846062B
- Authority
- CN
- China
- Prior art keywords
- user
- feature words
- user interest
- users
- interest model
- 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.)
- Active
Links
Classifications
-
- 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
- H04L67/55—Push-based network services
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to a kind of method for pushing based on users ' individualized requirement, including construct user interest model in server-side first；Then according to the user interest model, corresponding push result is returned to client；Wherein, the user interest model is to be created in the server-side according to current user interest feature set, and the user interest model further includes user identifier, and the user identifier is corresponding with the current user interest feature set.Using the method for pushing based on users ' individualized requirement in the invention, realize that user without the personalized push under deliberately operational hypothesis, has wider application range.
Description
Technical field
The present invention relates to field of computer technology more particularly to push technology fields, in particular to one kind based on user
The method for pushing of property demand.
Background technique
Currently, people usually by searched for after search engine inputs keyword information in the way of obtain the network information, but
It is that this mode hit rate based on keyword query is low, and it is difficult to meet different purposes, different background and different times
Inquiry request, while searching for data, tracking document respectively in multiple databases, have changed into a very cumbersome thing.
Summary of the invention
The purpose of the present invention is overcoming the above-mentioned prior art, a kind of personalized push that can be realized is provided
Method for pushing based on users ' individualized requirement.
To achieve the goals above, the method for pushing of the invention based on users ' individualized requirement has following constitute:
The method for pushing based on users ' individualized requirement, is mainly characterized by, the method the following steps are included:
(1) user interest model is constructed in server-side；
(2) according to the user interest model, corresponding push result is returned to client；
The user interest model is to be created in the server-side according to current user interest feature set, and described
User interest model further include user identifier, the user identifier is corresponding with the current user interest feature set.
In the method for pushing based on users ' individualized requirement,
The current user interest feature set is used by the Feature Words subset of user individual manual classification and early period
The set that Feature Words subset after the study of family interest model is constituted.
In initialization, Feature Words subset is the user interest profile collection of the method for pushing based on users ' individualized requirement
One empty set.
The Feature Words subset of the user individual manual classification of the method for pushing based on users ' individualized requirement has use
The hierarchical system of set membership in family personalization classification.
In the method for pushing based on users ' individualized requirement, the Feature Words subset of user individual manual classification be by with
The binary group collection that the weight of the Feature Words of family personalization manual classification and the Feature Words of the user individual manual classification is constituted,
Feature Words subset after the described user interest model early period study be by after the study of user interest model early period Feature Words and
The binary group collection that the weight of Feature Words after the study of user interest model early period is constituted.
In the method for pushing based on users ' individualized requirement, the weight of the Feature Words of the user individual manual classification
And the default-weight of the Feature Words after the study of user interest model early period is 1.
In the method for pushing based on users ' individualized requirement, the Feature Words subset of the user individual manual classification
The weight of corresponding parent Feature Words is calculated by following formula:
Wherein, f (t) is the power of parent Feature Words corresponding to the Feature Words subset of the user individual manual classification
Weight, c are the empirical value for being defaulted as 1, and the value of l is 0,1 or 2.
Using the method for pushing based on users ' individualized requirement in the invention, have the advantages that
(1) it does not need to preset user interest data, without cold beginning problem and Sparse Problems；
(2) can recommend for the user with particular interest preference；
(3) by list readers ' reading historical data, user interest model, document to be recommended content characteristic, can be with
It explains and recommends those documents why.
Detailed description of the invention
Fig. 1 is the application schematic diagram of the method for pushing of the invention based on users ' individualized requirement.
Specific embodiment
It is further to carry out combined with specific embodiments below in order to more clearly describe technology contents of the invention
Description.
The present invention relates to a kind of method for pushing based on users ' individualized requirement, comprising the following steps:
(1) user interest model is constructed in server-side；
(2) according to the user interest model, corresponding push result is returned to client；
The user interest model is to be created in the server-side according to current user interest feature set, and described
User interest model further include user identifier, the user identifier is corresponding with the current user interest feature set.
In the method for pushing based on users ' individualized requirement,
The current user interest feature set is used by the Feature Words subset of user individual manual classification and early period
The set that Feature Words subset after the study of family interest model is constituted.
In initialization, Feature Words subset is the user interest profile collection of the method for pushing based on users ' individualized requirement
One empty set.
The Feature Words subset of the user individual manual classification of the method for pushing based on users ' individualized requirement has use
The hierarchical system of set membership in family personalization classification.
In the method for pushing based on users ' individualized requirement, the Feature Words subset of user individual manual classification be by with
The binary group collection that the weight of the Feature Words of family personalization manual classification and the Feature Words of the user individual manual classification is constituted,
Feature Words subset after the described user interest model early period study be by after the study of user interest model early period Feature Words and
The binary group collection that the weight of Feature Words after the study of user interest model early period is constituted.
In the method for pushing based on users ' individualized requirement, the weight of the Feature Words of the user individual manual classification
And the default-weight of the Feature Words after the study of user interest model early period is 1.
In the method for pushing based on users ' individualized requirement, the Feature Words subset of the user individual manual classification
The weight of corresponding parent Feature Words is calculated by following formula:
Wherein, f (t) is the power of parent Feature Words corresponding to the Feature Words subset of the user individual manual classification
Weight, c are the empirical value for being defaulted as 1, and the value of l is 0,1 or 2.
Refering to Figure 1, it is based on for the method for pushing of the invention based on users ' individualized requirement is applied to one
Application schematic diagram in the one-stop search method for pushing of users ' individualized requirement, comprising:
(1) according to a user search request, the search source session in client is created；
(2) whether the judgement server-side supports the search source session, if supporting, continues step (3), otherwise
Error message is shown to the client and terminates whole process；
(3) search expression is obtained after being filtered according to filter expression to the session of described search source, if described search
Expression formula is invalid, then shows error message to the client and terminate whole process, otherwise continue step (4)；
(4) according to the search expression, and the user interest model obtained in the server-side, the clothes
Business end returns to corresponding one-stop search and push result to the client.
In the one-stop search method for pushing based on users ' individualized requirement, the step (1), including following step
It is rapid:
(1.1) it is created by the client and sends user search request；
(1.2) server-side described in receives the user search request, and after resolving to search expression, creation is simultaneously
Send search sessions；
(1.3) client described in receives the search sessions, and after having inquired session information, creates the meeting of search source
Words.
In the step of one-stop search method for pushing based on users ' individualized requirement (4), the search expression
It is made of unified structured expression, in order to match with the searching requirement of each database.
The structured expression of the one-stop search method for pushing based on users ' individualized requirement is searched for by the user
Logic unit, logical groups and the logic tree of demand are constituted.
In a specific embodiment, which includes logic unit, logical groups and logic tree:
(1) logic unit: the smallest querying condition unit, table 1 show field name in logic unit and corresponding
Illustrate:
Table 1
| Field name | Explanation of field |
| logic | Logical connector: " and ", "or", " non-" |
| field | Inquiry word segment limit |
| token | Relational operator |
| value | Field Inquiry value |
(2) logical groups: multiple logic units are combined, such as: " age is greater than 18 and height is greater than 170 ", table 2 shows
Go out the field name and respective description in logical groups:
Table 2
| Field name | Explanation of field |
| logic | Logical connector: " and ", "or", " non-" |
| units | The logic unit that current group includes |
(3) logic tree: the logic tree formed as unit of logical groups can express advanced search, and table 3 is shown in logic tree
Field name and respective description:
Table 3
| Field name | Explanation of field |
| logic | Logical connector: " and ", "or", " non-" |
| units | The logic unit that current group includes |
| children | The child node that current group includes |
Table 4
In a specific embodiment, it is somebody's turn to do the search source meeting of the one-stop search method for pushing based on users ' individualized requirement
It talks about shown in structure table 4 as above.
In a specific embodiment, should be based in the one-stop search method for pushing of users ' individualized requirement, it will be described
Search expression, and the combined value of the user interest model that is obtained in the server-side as inquiry session, wherein main
It acts on to retain search type, the structure for inquiring conversational list is as shown in table 5 below:
Table 5
| Field name | Field type | Explanation of field |
| key | integer | Session id |
| query | varcharacter(4096) | Query expression |
| createTime | timestamp | The conversation establishing time |
| account | integer | Inquire the ID of user |
Table 6
It in a specific embodiment, should be described based in the one-stop search method for pushing of users ' individualized requirement
During server-side returns to corresponding one-stop search result to the client, need that services sites is selected to be looked into
It askes, wherein main function is to retain search context property, as the inquiry of original database manufacturer needs Cookie or inquiry meeting
Information, the upper tables 6 such as words ID show website session table structure.
In a specific embodiment, user interest profile of the invention is concentrated: the feature of user individual manual classification
Word and the Feature Words after user interest model study are all the set of a binary group, have respectively included Feature Words
And the corresponding weight of Feature Words.Wherein, Feature Words of user individual manual classification and through the user interest model
The weight of Feature Words after habit, default are set as 1；It is worth noting that, the parent of the Feature Words of user individual manual classification is special
The weight f (t) for levying word, is calculated using formula one:
In above-mentioned formula one, l is characterized the number of plies that word differs in classification system structure, and the empirical value of c is set as 1.
In a specific embodiment, in the method for pushing of the invention based on users ' individualized requirement, user interest mould
Type is indicated with a binary group: I=(N, W), and wherein N indicates the corresponding reader's mark of user interest model, and W indicates reader's interest
Feature.W=(Wu, Wp, Wi) indicates the corresponding label word of this model；Wu indicates the document feature word by reader's manual classification；
The parent Feature Words of Wp expression Wu；Wi is indicated in the Feature Words (text crossed by readers ' reading increased newly after i study and adjustment
Offer the Feature Words of middle extraction), in initial user interest model I (0), there are no being refreshed using feedback information to it,
So initial Wi is an empty set.
In a specific embodiment, push of the invention is the result is that based on vector space model.Vector space model
A vector is expressed as with characteristic item and its corresponding weight value come the semanteme of characterization information, new information and user interest model.?
When needing to push new information to reader, the degree of correlation of new information and user interest model is described by vector operation.Make
Their similarity is calculated with the included angle cosine between vector.
In a specific embodiment, the data list structure in the method for pushing of the invention based on users ' individualized requirement
With reader's document category as shown in table 7:
Table 7
In a specific embodiment, the data list structure in the method for pushing of the invention based on users ' individualized requirement
With document category as shown in table 8 and read document mapping:
Table 8
In a specific embodiment, the data list structure in the method for pushing of the invention based on users ' individualized requirement
With document table as shown in table 9:
Table 9
Using the method for pushing based on users ' individualized requirement in the invention, have the advantages that
(1) it does not need to preset user interest data, without cold beginning problem and Sparse Problems；
(2) can recommend for the user with particular interest preference；
(3) by list readers ' reading historical data, user interest model, document to be recommended content characteristic, can be with
It explains and recommends those documents why.
In this description, the present invention is described with reference to its specific embodiment.But it is clear that can still make
Various modifications and alterations are without departing from the spirit and scope of the invention.Therefore, the description and the appended drawings should be considered as illustrative
And not restrictive.
Claims (3)
1. a kind of method for pushing based on users ' individualized requirement, which is characterized in that the method the following steps are included:
(1) user interest model is constructed in server-side；
(2) according to the user interest model, corresponding push result is returned to client；
The user interest model is creates according to current user interest feature set in the server-side, and the use
Family interest model further includes user identifier, and the user identifier is corresponding with the current user interest feature set；
The current user interest feature set be by user individual manual classification Feature Words subset and early period user it is emerging
The set that Feature Words subset after interesting model learning is constituted；
The Feature Words subset of the user individual manual classification has the grade body of set membership in user individual classification
System；
The Feature Words subset of the user individual manual classification is by the Feature Words of user individual manual classification and described
The binary group collection that the weight of the Feature Words of user individual manual classification is constituted, after the user interest model early period study
Feature Words subset is by the Feature Words after the study of user interest model early period and the Feature Words after the study of user interest model early period
Weight constitute binary group collection；
The weight of parent Feature Words corresponding to the Feature Words subset of the user individual manual classification is by following formula meter
It obtains:
Wherein, f (t) is the weight of parent Feature Words corresponding to the Feature Words subset of the user individual manual classification, c
For the empirical value for being defaulted as 1, l is characterized the number of plies that word differs in classification system structure, and the value of l is 0,1 or 2.
2. the method for pushing according to claim 1 based on users ' individualized requirement, which is characterized in that the user is emerging
For interesting feature set in initialization, Feature Words subset is an empty set.
3. the method for pushing according to claim 1 based on users ' individualized requirement, which is characterized in that the user personality
The default-weight of the weight and the Feature Words after the study of user interest model early period of changing the Feature Words of manual classification is 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810562061.5A CN108846062B (en) | 2018-06-04 | 2018-06-04 | Method for pushing based on users ' individualized requirement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810562061.5A CN108846062B (en) | 2018-06-04 | 2018-06-04 | Method for pushing based on users ' individualized requirement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108846062A CN108846062A (en) | 2018-11-20 |
| CN108846062B true CN108846062B (en) | 2019-08-13 |
Family
ID=64210619
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810562061.5A Active CN108846062B (en) | 2018-06-04 | 2018-06-04 | Method for pushing based on users ' individualized requirement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108846062B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101923545A (en) * | 2009-06-15 | 2010-12-22 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN101937524A (en) * | 2009-06-30 | 2011-01-05 | 华中师范大学 | Graduation design personalized guide system |
| CN102831199A (en) * | 2012-08-07 | 2012-12-19 | 北京奇虎科技有限公司 | Method and device for establishing interest model |
| CN104298732A (en) * | 2014-09-29 | 2015-01-21 | 中国科学院计算技术研究所 | Personalized text sequencing and recommending method for network users |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6981040B1 (en) * | 1999-12-28 | 2005-12-27 | Utopy, Inc. | Automatic, personalized online information and product services |
| CN101339562A (en) * | 2008-08-15 | 2009-01-07 | 北京航空航天大学 | Portal personalized recommendation service system introducing into interest model feedback and update mechanism |
| CN103455485A (en) * | 2012-05-28 | 2013-12-18 | 中兴通讯股份有限公司 | Method and device for automatically updating user interest model |
| CN107958070B (en) * | 2017-12-05 | 2021-11-12 | 上海电机学院 | Personalized message pushing method based on user preference |
-
2018
- 2018-06-04 CN CN201810562061.5A patent/CN108846062B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101923545A (en) * | 2009-06-15 | 2010-12-22 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN101937524A (en) * | 2009-06-30 | 2011-01-05 | 华中师范大学 | Graduation design personalized guide system |
| CN102831199A (en) * | 2012-08-07 | 2012-12-19 | 北京奇虎科技有限公司 | Method and device for establishing interest model |
| CN104298732A (en) * | 2014-09-29 | 2015-01-21 | 中国科学院计算技术研究所 | Personalized text sequencing and recommending method for network users |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108846062A (en) | 2018-11-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106776711B (en) | Chinese medical knowledge map construction method based on deep learning | |
| Zhao et al. | Connecting social media to e-commerce: Cold-start product recommendation using microblogging information | |
| Krishnapuram et al. | Low-complexity fuzzy relational clustering algorithms for web mining | |
| US20150262069A1 (en) | Automatic topic and interest based content recommendation system for mobile devices | |
| US7761464B2 (en) | Diversifying search results for improved search and personalization | |
| CN101216853B (en) | Intelligent web enquiry interface system and its method | |
| CN107590128B (en) | Paper homonymy author disambiguation method based on high-confidence characteristic attribute hierarchical clustering method | |
| EP1708099A1 (en) | Schema matching | |
| CN107391687A (en) | A kind of mixing commending system towards local chronicle website | |
| CN101496003A (en) | Compatibility scoring of users in a social network | |
| CN104834679A (en) | Representation and inquiry method of behavior track and device therefor | |
| CN106250545A (en) | A kind of multimedia recommendation method and system searching for content based on user | |
| Zhou et al. | Real-time context-aware social media recommendation | |
| Kwapong et al. | A knowledge graph based framework for web API recommendation | |
| CN113268667B (en) | Chinese comment emotion guidance-based sequence recommendation method and system | |
| CN105787066A (en) | Digital content distribution system based on total analysis | |
| CN106777359A (en) | A kind of text services based on limited Boltzmann machine recommend method | |
| Yang et al. | Design and application of handicraft recommendation system based on improved hybrid algorithm | |
| Nasery et al. | Polimovie: a feature-based dataset for recommender systems | |
| CN108846062B (en) | Method for pushing based on users ' individualized requirement | |
| Yu et al. | Web items recommendation based on multi-view clustering | |
| Nirwan et al. | Personalized hybrid book recommender system using neural network | |
| Sidana | Recommendation systems for online advertising | |
| CN108900574A (en) | One-stop search method for pushing based on users ' individualized requirement | |
| Silva et al. | HSLD: a hybrid similarity measure for linked data resources |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Fu Chen Inventor after: Lin Weixiao Inventor after: Xia Tian Inventor after: Dao Li Inventor after: Zhang Cheng Inventor after: Xia Han Inventor after: Liu Xinghang Inventor after: Mao Dan Inventor after: Cai Renzhi Inventor after: Geng Yibing Inventor before: Fu Chen Inventor before: Lin Weixiao Inventor before: Xia Tian Inventor before: Xia Han Inventor before: Zhang Cheng Inventor before: Dao Li Inventor before: Liu Xinghang Inventor before: Mao Dan Inventor before: Cai Renzhi Inventor before: Geng Yibing |