CN107562939A - Vertical field news recommends method, apparatus and readable storage medium - Google Patents
Vertical field news recommends method, apparatus and readable storage medium Download PDFInfo
- Publication number
- CN107562939A CN107562939A CN201710862705.8A CN201710862705A CN107562939A CN 107562939 A CN107562939 A CN 107562939A CN 201710862705 A CN201710862705 A CN 201710862705A CN 107562939 A CN107562939 A CN 107562939A
- Authority
- CN
- China
- Prior art keywords
- news
- user
- website
- data
- interest module
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a kind of Website News to recommend method, apparatus and readable storage medium storing program for executing.The Website News recommend method to comprise the following steps:User interest module is established according to the user characteristic data that portal website obtains;News template is established according to the news features data that portal website stores;Recommended by presetting recommendation rules with reference to the news that vertical field is carried out with interest module with news template, and the news of recommendation is shown.The present invention is by establishing line module and news template, so as to accurately obtain the interest module of user and the attribute of news, hence in so that the news in vertical field recommends accuracy to be greatly improved with professional so that user is greatly increased for the viscosity of portal website, and usage experience is also lifted.
Description
Technical field
The present invention relates to data mining and machine learning field, more particularly to a kind of Website News recommend method, apparatus and
Readable storage medium.
Background technology
All using the function of recommending for news, current news recommends method general for current most of portal website
It is that news corresponding to the information such as the keyword for the news or search clicked on according to user are carried out is recommended.Such news is recommended
Although method can quickly recommend news largely relevant with keyword, it can not be directed to and the focus of user is carried out deeply
The excavation entered, although user can obtain substantial amounts of news and recommend, however simultaneously also can be because the news content of recommendation can not be passed through
It is accurate to obtain the content for wanting concern, and lose the trust to website, ultimately result in the bad shadows such as attention rate reduction to website
Ring.
The content of the invention
It is a primary object of the present invention to provide a kind of Website News to recommend method, it is intended to which solving portal website can not be accurate right
The news that user accurately carries out vertical field is recommended.
To achieve the above object, the present invention provides a kind of Website News recommendation method, and the Website News recommend method bag
Include following steps:
User interest module is established according to the user characteristic data that website obtains;
News template is established according to the news features data that website stores;
According to default recommendation rules combination user interest module and news template, the recommendation news in vertical field is generated, and
It is shown recommending news to send to user terminal.
The step of user characteristic data obtained according to website establishes user interest module includes:
The essential information data of user in the user characteristic data are obtained, are passed through when user terminal browses web sites cold start-up
The first interest module that essential information data are established, wherein user interest module include the first interest module.
Include after the step of user interest module when establishing cold start-up by essential information data:
Browsing history based on user terminal obtains user's classification business development information, analysis user's classification business development
For information to obtain the second interest module on the short-term preference of user, wherein user interest module also includes the second interest module.
Include after the step of analysis user classifies business development information to obtain the second interest module:
The behavioral data of user is obtained, analyzes user behavior data to obtain the 3rd interest mould on the long-term preference of user
Block, wherein user interest module also include the 3rd interest module.
Alternatively, the step of news features data stored according to website establish news template includes:
The news data of the textual form of website storage is obtained, and data structure processing is carried out to news data to generate number
According to the news data of change;
The news template of digitization is established according to the digitization news data after processing.
It is described that news data progress data structure processing is included with generating the step of the news data of digitization:
The news data of textual form is converted into corresponding keyword vector, and digitization is drawn according to keyword vector
News data.
Alternatively, it is described by default recommendation rules combination user interest module and news template, generate vertical field
The step of recommending news includes:
Primary recommendation news is generated after being recommended according to the recommendation rules and user interest module and news template, just
Level recommends peg of news expert opinion further to be screened, and ultimately produces recommendation news.
Alternatively, it is described that news will be recommended to send to user terminal the step of being shown and include:
The recommendation news got is subjected to global alignment and is shown in client.
In addition, to achieve the above object, the present invention also provides a kind of news recommendation apparatus, the news recommendation apparatus bag
Include:Memory, processor and the news recommended program that can be run on the memory and on the processor is stored in, it is described
The step of Website News as described above recommend method is realized when news recommended program is by the computing device.
In addition, to achieve the above object, the present invention also provides a kind of computer-readable recording medium, described computer-readable
News recommended program is stored with storage medium, the news recommended program realizes news as described above when being executed by processor
The step of recommendation method.
Website News proposed by the present invention recommend method, and readding for user is counted and obtain by establishing user interest module
Custom, interest and behavior are read, and analyzes the news type that user wants to obtain.Again news is obtained by establishing news template
Correlation type, and by carrying out the matching of certain rule with user interest, draw the news currently recommended.The present invention can to
Comprehensive news reads tracking to family brachymedial for a long time, and accurately obtains the news reading requirement of user, and can be accurate
The news of true carry out association area is recommended, and accuracy is effectively ensured with professional, so that user is in portal website
The news that high quality can be got is recommended, and improves usage experience of the user for portal website.
Brief description of the drawings
Fig. 1 be the hardware running environment that scheme of the embodiment of the present invention is related to terminal apparatus structure schematic diagram;
Fig. 2 is the schematic flow sheet that Website News of the present invention recommend the embodiment of method one;
Fig. 3 is the refinement schematic flow sheet for the step of Website News of the present invention recommend S10 in another embodiment of method;
Fig. 4 is that Website News of the present invention recommend method modular structure schematic diagram;
Fig. 5 is that Website News of the present invention recommend method workflow diagram schematic diagram.
The realization, functional characteristics and advantage of the object of the invention will be described further referring to the drawings in conjunction with the embodiments.
Embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
As shown in figure 1, Fig. 1 is the terminal structure schematic diagram for the hardware running environment that scheme of the embodiment of the present invention is related to.
Terminal of the embodiment of the present invention can be PC or smart mobile phone, tablet personal computer, E-book reader, MP3
(Moving Picture Experts Group Audio Layer III, dynamic image expert's compression standard audio aspect 3)
Player, MP4 (Moving Picture Experts Group Audio Layer IV, dynamic image expert's compression standard sound
Frequency aspect 3) player, pocket computer etc. have the packaged type terminal device of display function.
As shown in figure 1, the terminal can include:Processor 1001, such as CPU, network interface 1004, user interface
1003, memory 1005, communication bus 1002.Wherein, communication bus 1002 is used to realize the connection communication between these components.
User interface 1003 can include display screen (Display), input block such as keyboard (Keyboard), optional user interface
1003 can also include wireline interface, the wave point of standard.Network interface 1004 can optionally connect including the wired of standard
Mouth, wave point (such as WI-FI interfaces).Memory 1005 can be high-speed RAM memory or stable memory
(non-volatile memory), such as magnetic disk storage.Memory 1005 optionally can also be independently of aforementioned processor
1001 storage device.
Alternatively, terminal can also include camera, RF (Radio Frequency, radio frequency) circuit, sensor, audio
Circuit, WiFi module etc..Wherein, sensor ratio such as optical sensor, motion sensor and other sensors.Specifically, light
Sensor may include ambient light sensor and proximity transducer, wherein, ambient light sensor can according to the light and shade of ambient light come
The brightness of display screen is adjusted, proximity transducer can close display screen and/or backlight when mobile terminal is moved in one's ear.As
One kind of motion sensor, gravity accelerometer can detect in all directions the size of (generally three axles) acceleration, quiet
Size and the direction of gravity are can detect that when only, the application available for identification mobile terminal posture is (such as horizontal/vertical screen switching, related
Game, magnetometer pose calibrating), Vibration identification correlation function (such as pedometer, tap) etc.;Certainly, mobile terminal can also match somebody with somebody
The other sensors such as gyroscope, barometer, hygrometer, thermometer, infrared ray sensor are put, will not be repeated here.
It will be understood by those skilled in the art that the restriction of the terminal structure shown in Fig. 1 not structure paired terminal, can be wrapped
Include than illustrating more or less parts, either combine some parts or different parts arrangement.
As shown in figure 1, it can lead to as in a kind of memory 1005 of computer-readable storage medium including operating system, network
Believe module, Subscriber Interface Module SIM and Website News recommended program.
In the terminal shown in Fig. 1, network interface 1004 is mainly used in connecting background server, is carried out with background server
Data communicate;User interface 1003 is mainly used in connecting client (user terminal), enters row data communication with client;And processor
1001 can be used for calling the Website News recommended program stored in memory 1005, and perform following operate:
User interest module is established according to the user characteristic data that website obtains;
News template is established according to the news features data that website stores;
According to default recommendation rules combination user interest module and news template, the recommendation news in vertical field is generated, and
It is shown recommending news to send to user terminal.
Further, processor 1001 can call the Website News recommended program stored in memory 1005, also perform
Operate below:
The essential information data of user in the user characteristic data are obtained, are passed through when user terminal browses web sites cold start-up
The first interest module that essential information data are established, wherein user interest module include the first interest module.
Reference picture 2, first embodiment of the invention provide a kind of news and recommend method, and the Website News recommend method bag
Include:
Step S10, the interest module of line module is established according to user characteristic data;
Step S20, news template is established according to news features data;
Step S30, according to default recommendation rules combination user interest module and news template, generate the recommendation in vertical field
News, and be shown recommending news to send to user terminal.
Specifically, news commending system needs to establish user interest module first according to user characteristic data, to establish use
The interest module at family, wherein user characteristic data include, essential information, user behavior feature and classification operation expanding feature.So
Afterwards according to the characteristic of news, the news stored in the form of text on internet is converted into the data mode of structuring, with
Facilitate follow-up calculating.Finally the data of line module and news template are recommended by the generation of default news recommendation rules new
Hear, and send to user terminal (browser or client).
Current most of portal websites etc. all can carry out news recommendation according to the use habit of user, and current news pushes away
Recommend method and be generally and noted down according to the click of user or the data such as search record recommend associated news that (such as user is led to
Search Liu De China is crossed to obtain the news related to Liu De China, then portal website can increase the news related to Liu De China, even
It is the recommendation of entertainment newses).This news recommendation method perhaps of short duration can provide the user with the facility of acquisition of information, but
Not real intelligent recommendation.User obtains certain news by the mode such as searching for or clicking on, it may be possible to because work
Or the personal short time needs, and and impersonal hobby, therefore to relevant information without the demand used after, door
The related news that website continues to recommend do not have good effect not only, can led on the contrary for a user then such as junk information
Apply the dislike at family.
If it is intended to increase user's viscosity, then be required to accurately find out the demand point of user, the present invention is by new
The technologies such as the keyword extraction of news, motif discovery are modeled to news, and non-structured data are converted into storable knot
Structure data.Then constructed by the analysis to user basic information and behavior and update line module.Recommended again by mixing
Algorithm, with reference to the expertise of industry, provide the user needs or information interested.It is anti-to recommendation list finally by user
The analysis of feedback, list constituent is adjusted, recommendation list is more personalized with intelligentized purpose when reaching.
What it is firstly the need of foundation is line module, solves the difference in recommendation process by getting user characteristic data
The problem of situation faces.User characteristic data includes essential information, user behavior feature and classification operation expanding feature.Substantially
The information that information is filled in by user when website is registered, carry out during to new user's cold start-up without reading histories simple
News is recommended;User behavior feature is mainly user's reading histories and user's reading operations behavior, by the reading for analyzing user
Operation behavior data more can accurately obtain the interest and preference of user, the system can by the keyword in related news or
The weight of name entity is lifted, and is stored in the Long-term Interest module of user.User Part operation behavior and mutually speak on somebody's behalf
It is bright to introduce such as table 1.
Table 1
And operation expanding information of classifying is class of service (herein referring to the news that the user read) extraction used from user
The user characteristics for expressing user preference.With the increase of user's news amount of reading, extension information can also increase therewith, therefore
Need in time to be updated it, accurately to catch user's short-term interest.
The feature of three aspect users is finally integrated, carrying out comprehensive modeling to user portrays user.User i module
UiFor U={ Ii, RHi, Pi}.Wherein IiRepresent user i essential information, Ii={ gender=0, age=4... };RHiRepresent
User i reads the ID of the article of most worthy, takes i=20, RHI={NT1, NT2... NT20};PiRepresent user i behaviors spy
The keyword that can represent user section period interest that sign obtains jointly with classification operation expanding information.Pi={ ki1,
ki2... ki20}。
In addition to establishing user interest module, the present invention also needs to establish news news, so as to according to the emerging of user
News corresponding to interest progress is recommended.News stores in a text form on the internet, belongs to unstructured data.To new
Need to handle work by a series of data structured when news is modeled, can store, count so as to which text data be converted to
The structural data of calculation.News features are divided into three classes herein:Essential characteristic, text feature and name entity.By news
Modeling process, the news that an ID is j may be expressed as Nj={ NIj, Kj}.Wherein NIjRepresent news j essential characteristic, KjGeneration
What table extracted from news j can replace the set of the text feature and name entity of news.
Essential characteristic NIj have recorded the most basic feature of news, and major part carries attribute when being news briefing.Substantially it is special
Sign designs program when news captures and captured, and is stored in the NIj in news template, is asked available for solving cold start-up
Topic.News essential characteristic list such as table 2.
Table 2
In addition to essential characteristic, news also has text feature, text feature be from the text extracting of news to can be with
Show a series of keywords (or keyword) of theme of news.Name entity be intrinsic title in text, abbreviation and other only
One mark, generally includes 7 kinds of classifications:Personage, mechanism, place, date, time, money and percentage.In different portal websites
In, the Chinese name entity storehouse of peculiar vocabulary is often created as needed, is named in fact rapidly and accurately to be identified in news
Body.Then according to the frequency occurred according to keyword in news, (keyword positioned at the position such as title or subtitle exists for position
Express news content when, have more importantly status) etc. factor determine weights of the keyword in Present News.
The proposed algorithm of the present invention employs mixing proposed algorithm, i.e., according to the reading habit of user, recommends to calculate with three kinds
Method (AJS proposed algorithms, AK-means proposed algorithms, ABC-BC proposed algorithms, being the common algorithms of current data mining) gained
As a result merged.Wherein, every kind of proposed algorithm is improved according to the characteristics of vertical door with demand, more accurate to obtain
Recommendation results (present system workflow diagram such as Fig. 5).Three kinds of recommendation methods are respectively content-based recommendation, association rule
Then and collaborative filtering, it is classic algorithm to use.This patent highlights the design of each module for vertical door, and each
Relation and structure (modular structure schematic diagram such as Fig. 4) between individual module.
Website News of the present invention recommend method in order to allow recommendation list more rationally and hommization, for recommending by news
Some news () that system selects, system can be ranked up according to the news briefing time.Sorted by news so that when
The strong news of effect property possesses the chance more read, and makes recommendation list more reasonable, more readable.And can also be by compiling
Collect (management staff) and manually recommend and obtain the mode such as number of clicks highest news (i.e. hot news) in certain time
Carry out three-dimensional news to recommend, so as to ensure that user can get most accurate news in time.
Further, reference picture 3, state the step of user characteristic data obtained according to website establishes user interest module and wrap
Include:
Step S11, the essential information data of user in the user characteristic data are obtained, browsed web sites cold open in user terminal
The the first interest module established when dynamic by essential information data, wherein user interest module include the first interest module.
Specifically, the essential information that user registers in website is obtained, and the first interest of user is established according to essential information
Module, and the first interest module is mainly used in carrying out news recommendation during user's cold start-up.
User can be required one account of registration first when using portal website, and when Account Registration is carried out,
Then need to fill in some essential informations to help through the registration of account.It is of the invention then be by obtain register when basic letter
Breath, news during solving user's cold start-up recommend that (i.e. user browses portal website first, or clear during without reading histories
Look at), during user's cold start-up because user does not have reading histories etc. to be analyzed and be referred to, thus when recommending news according to
The Back ground Information of user, recommend the news relevant with data in Back ground Information, at the same find out with user have similar interests its
He is user, is recommended according to the preference of the other users similar to user interest.Registration required by different portal websites
Information page differs, and is often adjusted accordingly according to the content of each portal website, such as campus network may require that and fill in student number
Information such as () specialty of user can be obtained, whether graduated, and sports portal then may require that to fill in and be good at or interested
Sports items (can get user's sports items interested).Therefore when user carries out cold start-up (except Back ground Information
Outside, without other data such as reading histories of user), user interest mould can be established according to Back ground Information during user's registration
Block, to carry out related news recommendation.
Further, wrapped after the step of user interest module when establishing cold start-up by essential information data
Include:
Step S12, the browsing history based on user terminal obtain user's classification business development information, analysis user's classification
For business development information to obtain the second interest module on the short-term preference of user, it is emerging also to include second for wherein user interest module
Interesting module.
Specifically, by obtaining the classification operation expanding information data of user and analyzing, so as to establish the short-term preference of user
The second interest module on the short-term preference of user.
Classification operation expanding information is that the news that certain amount is read recently is extracted from the news experience history of user, is passed through
The news reading histories are analyzed, the second interest module of the short-term preference of user more can be accurately obtained with acquisition.In order to
Guaranteeing accurately to obtain user's short-term interest, the sample that classification business development information is extracted is also required to upgrade in time, and
And with the increase of user's news amount of reading, expand information and be also required to increase, to ensure the accuracy of user characteristics, so as to
The short-term interest module of user can accurately be established.
News has actual effect, and user may increase to some in the short term because of reasons such as work or individuals
Or the attention rate of the news content in some fields, such as during world cup, user can be for the reading of the related news of football
Frequency may be significantly increased, then show that user shows larger interest for world cup, then system can increase in a short time
Add and push the news related to world cup on football, to meet user in a short time for football and world cup related news
Demand.By obtaining gangster's class business development information data of user, the interest module that the short-term news of user is read is established, so as to
More accurately obtain the short-term interest of user.
Further, wrapped after the step of analysis user classifies business development information to obtain the second interest module
Include:
Step S13, obtains the behavioral data of user, and analysis user behavior data is to obtain the on the long-term preference of user
Three interest modules, wherein user interest module also include the 3rd interest module.
Specifically, the behavioral data according to the user got in reading histories and reading operations behavior, and according to right
The analysis of user behavior data obtains the 3rd interest module of the long-term preference of user.
For user when reading news, meeting is different from interest according to reading purpose and carries out different reading operations, according to
The operation behavior for reading news can be to user reading preference carry out analysis judgement with reading purpose etc..Reading operations behavior refers to
Be user when carrying out news reading, a series of behaviour in addition to reading for being carried out according to the difference of interest or purpose
Make.Including collection, printing, comment etc., the meaning representated by different operation behaviors is also different, and can be by user
Had in the news of reading operation behavior news carry out weight lifting (in general, user can pair with it is interested new
News is operated, and be only the news wanted to know about then operate it is relatively fewer), therefore can be according to user's associative operation behavior
Number of operations establish the 3rd interest module of more detailed long-term preference.
And the long-term interest module of user is established, the preference of user can be more accurately analyzed, helps user to be easier
The deep content in interested some or some fields is obtained, and user is as depth and high quality can have been got
News so that user establishes higher adhesive to being used caused by portal website, so as to improve official loyal to his sovereign's degree of user.
Further, the step of news features data stored according to website establish news template includes:
Step S21, the news data of the textual form of website storage is obtained, and data structure processing is carried out to news data
To generate the news data of digitization;
Step S22, the news template of digitization is established according to the digitization news data after processing.
Specifically, news is stored in a text form on network, and the form of text is not used to establish
News template, it is therefore desirable to the news content of textual form is converted into the processing of data structure, and according to obtained data knot
The news template of the vertical digitization of structure.
User is to be stored in a text form in Website server in the news that portal website reads, and text
The news of form needs to handle by the structuring of data volume, to obtain the structured data of digitization, to be converted into the structure number of digitization
According to for be stored in calculate then more it is convenient (storage in a text form needs to waste substantial amounts of server space, and for
It is difficult to be directly used in calculating), so as to preferably establish the news template of digitization.Pass through building for the news template of digitization
It is vertical, the matching of recommendation rules is carried out with user interest module, then can get the news associated with user interest, and carry out
Recommend.
Further, it is described that news data progress data structure processing is wrapped with generating the step of the news data of digitization
Include:
Step S211, the news data of textual form is converted into corresponding keyword vector, and obtained according to keyword vector
Go out the news data of digitization.
Specifically, by the news of website by the data of textual form by extracting essential characteristic and text feature the methods of turn
Turn to the structured data of keyword vector form.
The news of portal website stores in the form of text, but when establishing news template, the news of textual form is nothing
Used between method, it is therefore desirable to thaumatropy is carried out, so as to preferably carry out the foundation of news template.By textual form
News carry out data structured processing be divided into two steps, be first obtain news essential characteristic.News essential characteristic have recorded newly
The most basic feature heard, major part carry attribute when being news briefing.News essential characteristic table such as table 3.
Table 3
The information such as the type of news can be got roughly by essential characteristic, it is quick so as to be carried out in cold start-up
News recommend.In addition to essential characteristic, it is also necessary to the identification and acquisition of text feature are carried out to news.The text of news is special
Sign be from the text extracting of news to a series of vocabulary that may indicate that theme of news, name entity is the proper name in text
Title, abbreviation and other unique marks, generally include 7 kinds of classifications:Personage, mechanism, place, date, time, money and percentage
Than.In different portal websites, the Chinese name entity storehouse of peculiar vocabulary is often created as needed, so as to quick in news
Identification name entity exactly.Keyword to make to select can more represent document main idea, typically use TF*IDF methods, formula is such as
Under:
The weights of each keyword are calculated, weighting value highest keyword is stored.Name entity, word are considered herein
There is different influences language position to news description, in order to make the keyword elected more accurately represent news content, this
Text is improved TF*IDF formula, and formula is as follows:
After Keyword Weight calculates completion, news template K is arrived in the 10 keywords storages of weighting value highestjIn.Wherein
WjkRepresent weights of the k keywords for the news that ID is j in this news;tfjkRepresent keyword kjkOccur in news j
Number;tdfkRepresent keyword kjkThe number occurred in all documents;W represents weights, and the system regulation occurs when keyword
In title, subtitle, keyword for name entity when, W be more than 1, otherwise W be less than or equal to 1 (positioned at title word and
Entity is named to have more importantly status, therefore weights are big when expressing news content).The improved foundation of formula is, positioned at mark
The word and name entity of topic have more importantly status, specific weights value is according to different when expressing news content
Depending on system.Modeled by news, the news vector module N that ID is j can be drawnj={ NIj, Kj}.Wherein NIjRepresent news j
Essential characteristic, KjRepresent the set of the text feature and name entity that can replace news that are extracted from news j.
Further, it is described by default recommendation rules combination user interest module and news template, generate vertical field
Recommendation news the step of include:
Step S31, primary is generated after being recommended according to the recommendation rules with user interest module and news template and is pushed away
News is recommended, primary recommends peg of news expert opinion further to be screened, ultimately produces recommendation news.
Specifically, primary recommendation news can be generated according to the recommendation rules and user interest module and news template,
And in order to improve the depth of the accuracy and content of recommending news, the present invention can also be according to the opinion pair of relevant industries professional
Primary news is screened, and obtains recommending news after screening.
The contents such as news, therefore vertical door position are provided because vertical gate family is mainly directed towards a certain specific area (region)
Recommendation news generally comprise the related contents of some professional knowledges.And in order to ensure recommending the professional of news, meeting of the invention
The correlation rule formulated according to expert, by the potential contact between the keyword of relationship data mining, is obtained by association rule algorithm.
By the excavation of whole reading histories to all users, potential relation between keyword is obtained, to improve algorithm, particularly
The precision of content-based recommendation algorithm.By the expertise of specific area, as the background knowledge of commending system, can both solve
Cold start-up problem, the accuracy of recommendation can be improved again, improve degree of belief of the user to system.
Further, it is described that news will be recommended to send to user terminal the step of being shown and include:
Step S32, the recommendation news got is subjected to global alignment and is shown in user terminal.
Specifically, the news quantity got by recommendation rules is more, it is therefore desirable to is combined according to certain order
With arrangement, user terminal displaying is carried out at after arrangement.
News displaying list of the present invention mainly has three kinds of ways of recommendation, and system is adopted according to line module with reference to news template
Recommend news caused by mixing proposed algorithm;Recommend news according to caused by simple mapping ruler as editor (management staff)
And system in certain time according to according to clicking rate, choosing some news of highest, caused recommendation news.Pass through three kinds
The way of recommendation, more comprehensively it can show its news content for wanting to obtain for user, and by manually double with system
Recommend again, avoid the influence according only to human emotion or system-computed, caused by recommend news be deviated.And can also
Non- user recommends current most popular news, ensures no matter whether user is concerned about association area, most hot at present all without missing
The Domestic News of door.And the news emphasis that different recommendation rules are recommended is also different, can be sentenced according to line module
The emphasis of disconnected user, it is of the invention by by the news of recommendation so as to which preferably the news of recommendation is combined and arranged
It is combined with clapping so that user can get more comprehensive and accurate news content.
The present invention also provides a kind of device for recommending method based on Website News.
The present invention recommends the device of method to include based on Website News:Memory, processor and it is stored in the memory
News recommended program that is upper and can running on the processor, is realized when the news recommended program is by the computing device
News as described above recommends method and step.
Wherein, the method that the news recommended program run on the processor is realized when being performed can refer to the present invention
Website News recommend each embodiment of method, will not be repeated here.
In addition the embodiment of the present invention also proposes a kind of computer-readable recording medium.
News recommended program is stored with computer-readable recording medium of the present invention, the news recommended program is by processor
The step of news as described above recommends method is realized during execution.
Wherein, the method that the news recommended program run on the processor is realized when being performed can refer to the present invention
Website News recommend each embodiment of method, will not be repeated here.
It should be noted that herein, term " comprising ", "comprising" or its any other variant are intended to non-row
His property includes, so that process, method, article or system including a series of elements not only include those key elements, and
And also include the other element being not expressly set out, or also include for this process, method, article or system institute inherently
Key element.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that including this
Other identical element also be present in the process of key element, method, article or system.
The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can add the mode of required general hardware platform to realize by software, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on such understanding, technical scheme is substantially done to prior art in other words
Going out the part of contribution can be embodied in the form of software product, and the computer software product is stored in one as described above
In storage medium (such as ROM/RAM, magnetic disc, CD), including some instructions to cause a station terminal equipment (can be mobile phone,
Computer, server, air conditioner, or network equipment etc.) perform method described in each embodiment of the present invention.
The preferred embodiments of the present invention are these are only, are not intended to limit the scope of the invention, it is every to utilize this hair
The equivalent structure or equivalent flow conversion that bright specification and accompanying drawing content are made, or directly or indirectly it is used in other related skills
Art field, is included within the scope of the present invention.
Claims (10)
1. a kind of Website News recommend method, it is characterised in that the Website News recommend method to comprise the following steps:
User interest module is established according to the user characteristic data that website obtains;
News template is established according to the news features data that website stores;
According to default recommendation rules combination user interest module and news template, the recommendation news in vertical field is generated, and will be pushed away
Recommend news and send to user terminal and be shown.
2. Website News as claimed in claim 1 recommend method, it is characterised in that the user characteristics obtained according to website
Data, which establish the step of user interest module, to be included:
The essential information data of user in the user characteristic data are obtained, when user terminal browses web sites cold start-up by basic
The first interest module that information data is established, wherein user interest module include the first interest module.
3. Website News as claimed in claim 2 recommend method, it is characterised in that it is described established by essential information data it is cold
Include after the step of user interest module during startup:
Browsing history based on user terminal obtains user's classification business development information, analysis user's classification business development information
To obtain the second interest module on the short-term preference of user, wherein user interest module also includes the second interest module.
4. Website News as claimed in claim 3 recommend method, it is characterised in that the analysis user classification business development letter
Include after the step of breath is to obtain the second interest module:
The behavioral data of user is obtained, analyzes user behavior data to obtain the 3rd interest module on the long-term preference of user,
Wherein user interest module also includes the 3rd interest module.
5. Website News as claimed in claim 1 recommend method, it is characterised in that the news features stored according to website
Data, which establish the step of news template, to be included:
The news data of the textual form of website storage is obtained, and data structure processing is carried out to news data to generate digitization
News data;
The news template of digitization is established according to the digitization news data after processing.
6. Website News as claimed in claim 5 recommend method, it is characterised in that described to carry out data structure to news data
Processing is included with generating the step of the news data of digitization:
The news data of textual form is converted into corresponding keyword vector, and the news of digitization is drawn according to keyword vector
Data.
7. Website News as claimed in claim 1 recommend method, it is characterised in that described combined by default recommendation rules is used
Family interest module and news template, include the step of the recommendation news for generating vertical field:
Primary recommendation news is generated after being recommended according to the recommendation rules and user interest module and news template, primary pushes away
Recommend peg of news expert opinion further to be screened, ultimately produce recommendation news.
8. Website News as claimed in claim 1 recommend method, it is characterised in that described that news will be recommended to send to user terminal
The step of being shown includes:
The recommendation news got is subjected to global alignment and is shown in user terminal.
9. a kind of news recommendation apparatus, it is characterised in that described device includes:Memory, processor and it is stored in the storage
On device and the Website News recommended program that can run on the processor, the Website News recommended program is by the processor
The step of Website News as any one of claim 1 to 8 recommend method is realized during execution.
10. a kind of computer-readable recording medium, it is characterised in that it is new to be stored with website on the computer-readable recording medium
Recommended program is heard, is realized when the Website News recommended program is executed by processor as any one of claim 1 to 8
News recommends the step of method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710862705.8A CN107562939B (en) | 2017-09-21 | 2017-09-21 | Vertical domain news recommendation method and device and readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710862705.8A CN107562939B (en) | 2017-09-21 | 2017-09-21 | Vertical domain news recommendation method and device and readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107562939A true CN107562939A (en) | 2018-01-09 |
CN107562939B CN107562939B (en) | 2021-03-23 |
Family
ID=60982119
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710862705.8A Active CN107562939B (en) | 2017-09-21 | 2017-09-21 | Vertical domain news recommendation method and device and readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107562939B (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109300061A (en) * | 2018-08-31 | 2019-02-01 | 哈尔滨理工大学 | A kind of individual information recommendation plateform system based on deep learning tourism |
CN109660591A (en) * | 2018-11-02 | 2019-04-19 | 北京奇虎科技有限公司 | The automatic push method, apparatus and calculating equipment of Personalize News |
CN110413890A (en) * | 2019-07-29 | 2019-11-05 | 武汉匠楚科技有限公司 | A kind of method that news recommender system polymerization news is presented |
CN110580317A (en) * | 2019-08-29 | 2019-12-17 | 武汉赛可锐信息技术有限公司 | social information analysis method and device, terminal equipment and storage medium |
CN110688476A (en) * | 2019-09-23 | 2020-01-14 | 腾讯科技(北京)有限公司 | Text recommendation method and device based on artificial intelligence |
CN110866183A (en) * | 2019-11-06 | 2020-03-06 | 北京字节跳动网络技术有限公司 | Social interface recommendation method and device, electronic equipment and storage medium |
CN110968789A (en) * | 2019-12-04 | 2020-04-07 | 掌阅科技股份有限公司 | Electronic book pushing method, electronic equipment and computer storage medium |
CN111601164A (en) * | 2020-05-21 | 2020-08-28 | 广州欢网科技有限责任公司 | Intelligent television news pushing method and system |
CN111767466A (en) * | 2020-09-01 | 2020-10-13 | 腾讯科技(深圳)有限公司 | Recommendation information recommendation method and device based on artificial intelligence and electronic equipment |
CN114670760A (en) * | 2020-12-24 | 2022-06-28 | 九号智能(常州)科技有限公司 | Vehicle control method and device, electronic equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202131A (en) * | 2015-05-08 | 2016-12-07 | 蔡奇 | A kind of news based on user interest recommends method |
CN106383887A (en) * | 2016-09-22 | 2017-02-08 | 深圳市博安达信息技术股份有限公司 | Environment-friendly news data acquisition and recommendation display method and system |
US20170039275A1 (en) * | 2015-08-03 | 2017-02-09 | International Business Machines Corporation | Automated Article Summarization, Visualization and Analysis Using Cognitive Services |
CN107025310A (en) * | 2017-05-17 | 2017-08-08 | 长春嘉诚信息技术股份有限公司 | A kind of automatic news in real time recommends method |
-
2017
- 2017-09-21 CN CN201710862705.8A patent/CN107562939B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202131A (en) * | 2015-05-08 | 2016-12-07 | 蔡奇 | A kind of news based on user interest recommends method |
US20170039275A1 (en) * | 2015-08-03 | 2017-02-09 | International Business Machines Corporation | Automated Article Summarization, Visualization and Analysis Using Cognitive Services |
CN106383887A (en) * | 2016-09-22 | 2017-02-08 | 深圳市博安达信息技术股份有限公司 | Environment-friendly news data acquisition and recommendation display method and system |
CN107025310A (en) * | 2017-05-17 | 2017-08-08 | 长春嘉诚信息技术股份有限公司 | A kind of automatic news in real time recommends method |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109300061A (en) * | 2018-08-31 | 2019-02-01 | 哈尔滨理工大学 | A kind of individual information recommendation plateform system based on deep learning tourism |
CN109660591A (en) * | 2018-11-02 | 2019-04-19 | 北京奇虎科技有限公司 | The automatic push method, apparatus and calculating equipment of Personalize News |
CN110413890A (en) * | 2019-07-29 | 2019-11-05 | 武汉匠楚科技有限公司 | A kind of method that news recommender system polymerization news is presented |
CN110580317B (en) * | 2019-08-29 | 2022-02-22 | 武汉赛可锐信息技术有限公司 | Social information analysis method and device, terminal equipment and storage medium |
CN110580317A (en) * | 2019-08-29 | 2019-12-17 | 武汉赛可锐信息技术有限公司 | social information analysis method and device, terminal equipment and storage medium |
CN110688476A (en) * | 2019-09-23 | 2020-01-14 | 腾讯科技(北京)有限公司 | Text recommendation method and device based on artificial intelligence |
CN110866183A (en) * | 2019-11-06 | 2020-03-06 | 北京字节跳动网络技术有限公司 | Social interface recommendation method and device, electronic equipment and storage medium |
CN110866183B (en) * | 2019-11-06 | 2023-06-13 | 北京字节跳动网络技术有限公司 | Social interface recommendation method and device, electronic equipment and storage medium |
CN110968789A (en) * | 2019-12-04 | 2020-04-07 | 掌阅科技股份有限公司 | Electronic book pushing method, electronic equipment and computer storage medium |
CN110968789B (en) * | 2019-12-04 | 2023-05-23 | 掌阅科技股份有限公司 | Electronic book pushing method, electronic equipment and computer storage medium |
CN111601164A (en) * | 2020-05-21 | 2020-08-28 | 广州欢网科技有限责任公司 | Intelligent television news pushing method and system |
CN111767466A (en) * | 2020-09-01 | 2020-10-13 | 腾讯科技(深圳)有限公司 | Recommendation information recommendation method and device based on artificial intelligence and electronic equipment |
CN114670760A (en) * | 2020-12-24 | 2022-06-28 | 九号智能(常州)科技有限公司 | Vehicle control method and device, electronic equipment and storage medium |
CN114670760B (en) * | 2020-12-24 | 2023-08-18 | 九号智能(常州)科技有限公司 | Control method and device for vehicle, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN107562939B (en) | 2021-03-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107562939A (en) | Vertical field news recommends method, apparatus and readable storage medium | |
US7917514B2 (en) | Visual and multi-dimensional search | |
US7739221B2 (en) | Visual and multi-dimensional search | |
TWI533246B (en) | Method and system for discovery of user unknown interests | |
CN103455524B (en) | Method and device for displaying and acquiring entry information | |
US20130036121A1 (en) | System and method for recommending blog | |
CN110929021B (en) | Text information generation method and text information generation device | |
US20150154303A1 (en) | System and method for providing content recommendation service | |
CN108121736A (en) | A kind of descriptor determines the method for building up, device and electronic equipment of model | |
Strobbe et al. | Interest based selection of user generated content for rich communication services | |
CN111522940B (en) | Method and device for processing comment information | |
CN113569129A (en) | Click rate prediction model processing method, content recommendation method, device and equipment | |
CN110390569A (en) | A kind of content promotion method, device and storage medium | |
CN109582869A (en) | A kind of data processing method, device and the device for data processing | |
CN109101505A (en) | A kind of recommended method, recommendation apparatus and the device for recommendation | |
CN110347314A (en) | A kind of content displaying method, device, storage medium and computer equipment | |
KR102195686B1 (en) | Apparatus and method of recommending items based on areas | |
KR20110052114A (en) | Recommendation searching system using internet and method thereof | |
JP5805134B2 (en) | Terminal device and device program | |
CN114816180A (en) | Content browsing guiding method and device, electronic equipment and storage medium | |
JP6890764B2 (en) | Teacher data generation system, teacher data generation method, and program | |
CN116186413A (en) | Method and system for multi-modal feature recommendation | |
CN116205686A (en) | Method, device, equipment and storage medium for recommending multimedia resources | |
CN115168568A (en) | Data content identification method and device and storage medium | |
JP2021092931A (en) | Information processing device and program |
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 |