CN104361063B - user interest discovery method and device - Google Patents
user interest discovery method and device Download PDFInfo
- Publication number
- CN104361063B CN104361063B CN201410613040.3A CN201410613040A CN104361063B CN 104361063 B CN104361063 B CN 104361063B CN 201410613040 A CN201410613040 A CN 201410613040A CN 104361063 B CN104361063 B CN 104361063B
- Authority
- CN
- China
- Prior art keywords
- user
- interest
- data
- expression data
- content recommendation
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
- G06F16/337—Profile generation, learning or modification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The embodiment of the present invention, which provides a kind of user interest discovery method and device, this method, to be included：Obtain the expression data or behavioral data to content recommendation of user's input；According to the expression data or the behavioral data, interest prediction result is determined；The interest prediction result is prompted to applications client, and obtains click data of the user to the interest prediction result, to determine user interest.The embodiment of the present invention can allow for user actively to carry out interest expression by natural language, find user interest accurately and in time, lift Consumer's Experience.
Description
Technical field
The present embodiments relate to areas of information technology, more particularly to a kind of user interest discovery method and device.
Background technology
Personalized information recommending technology can issue the information for meeting user interest to user, and therefore, the technology gradually exists
More and more applied in network access., it is necessary to which discovery user accurately and timely is emerging in personalized information recommending technology
Interest.
Existing user interest discovery technique, click of the user to content recommendation is usually obtained, is shared, collection etc. is positive and negative
Feedback behavior and/or ignore, the negative-feedback behavior such as step on, and user interest is analyzed from feedback behavior.
Following defect be present in above-mentioned user interest discovery technique：On the one hand, the information of the feedback behavior reflection of user compares
It is fuzzy, a content recommendation, either positive feedback behavior or negative-feedback behavior are given, it is difficult to judge user feedback pin exactly
To specific object, the degree of accuracy of the user interest resulted in a finding that is low；On the other hand, a large amount of feedback behaviors of user are generally required
User interest could be accurately captured, because this process is generally time-consuming longer, therefore is unfavorable for finding user interest in time.
The content of the invention
The embodiment of the present invention provides a kind of user interest discovery method and device, to improve the accurate of the user interest found
Property and promptness.
The embodiment of the present invention uses following technical scheme：
On the one hand, the embodiments of the invention provide a kind of user interest discovery method, including：
Obtain the expression data or behavioral data to content recommendation of user's input；
According to the expression data or the behavioral data, interest prediction result is determined；
The interest prediction result is prompted to applications client, and obtains hits of the user to the interest prediction result
According to determine user interest.
Further, according to the expression data, interest prediction result is determined, including：
According to the expression data, it is determined that object of interest corresponding with the expression data and user's attitude；
According to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.
Further, before the expression data to content recommendation of user's input are obtained, in addition to：
Obtain the behavioral data to content recommendation of user's input；
According to the behavioral data, satisfaction of the user to the content recommendation is determined；
If the satisfaction is less than setting threshold value, triggering performs the expression to content recommendation for obtaining user's input
The operation of data.
Further, according to the expression data, it is determined that object of interest corresponding with the expression data, including：
The word that the expression data are included is matched in object repository；
Using the word that the match is successful as object of interest corresponding to the expression data.
Further, after using the word that the match is successful as object of interest corresponding to the expression data, in addition to：
According to the text distance between object of interest word corresponding with user's attitude, it is big to filter out text distance
In the object of interest of setting value.
Further, according to the expression data, it is determined that user's attitude corresponding with the expression data, including：
The word that the expression data are included is matched in default emotional attitude template；
According to matching result, it is determined that user's attitude corresponding with the expression data.
Further, according to the behavioral data, before determining interest prediction result, in addition to：
According to the behavioral data, satisfaction of the user to the content recommendation is determined；
If the satisfaction is less than setting threshold value, triggering is performed according to the behavioral data, determines that interest is predicted
As a result operation.
Further, according to the behavioral data, determine satisfaction of the user to the content recommendation, including it is following at least
One：
Refreshing frequency according to user to content recommendation, determine satisfaction of the user to the content recommendation；
Click data and stay time according to user to content recommendation, determine satisfaction of the user to the content recommendation
Degree；
According to support feedback data of the user to content recommendation and concern time, determine that user expires to the content recommendation
Meaning degree.
Further, the interest prediction result is being prompted to applications client, and is obtaining user and the interest is predicted
As a result click data, after determining user interest, in addition to：
According to the user interest of determination, the content recommendation pushed to user is corrected.
On the other hand, the embodiment of the present invention additionally provides a kind of user interest and finds device, including：
User data acquisition module, for obtaining the expression data or behavioral data to content recommendation of user's input；
Interest prediction result determining module, for according to the expression data or the behavioral data, determining that interest is predicted
As a result；
Interest determination module, for prompting the interest prediction result to applications client, and user is obtained to described emerging
The click data of interesting prediction result, to determine user interest.
Further, interest prediction result determining module includes：
Object of interest determining unit, for according to the expression data, it is determined that interest pair corresponding with the expression data
As；
User's attitude determining unit, for according to the expression data, it is determined that User space corresponding with the expression data
Degree；
Interest prediction result determining unit, for according to the corresponding object of interest of the expression data and user's attitude,
Determine interest prediction result.
Further, described device also includes：
First satisfaction determining module, for obtain user input the expression data to content recommendation before, according to
The behavioral data, determine satisfaction of the user to the content recommendation；
Data acquisition trigger module is expressed, for when the satisfaction is less than setting threshold value, triggering to be performed to obtain and used
The operation of the expression data to content recommendation of family input.
Further, object of interest determining unit includes：
Coupling subelement, the word included for matching the expression data in object repository；
Object of interest determination subelement, for using the word that the match is successful as object of interest corresponding to the expression data.
Further, object of interest determining unit also includes：
Object of interest filters subelement, for using the word that the match is successful as object of interest corresponding to the expression data
Afterwards, according to the text distance between object of interest word corresponding with user's attitude, filter out text distance and be more than
The object of interest of setting value.
Further, user's attitude determining unit is specifically used for：
The word that the expression data are included is matched in default emotional attitude template；
According to matching result, it is determined that user's attitude corresponding with the expression data.
Further, described device also includes：
Second satisfaction determining module, for according to the behavioral data, before determining interest prediction result, according to institute
Behavioral data is stated, determines satisfaction of the user to the content recommendation；
Interest prediction result determines trigger module, if being less than setting threshold value for the satisfaction, triggering performs
According to the behavioral data, interest prediction result is determined.
Further, the first satisfaction determining module or the second satisfaction determining module include at least one of following：
First satisfaction determining unit, for the refreshing frequency according to user to content recommendation, determine that user pushes away to described
Recommend the satisfaction of content；
Second satisfaction determining unit, for the click data and stay time according to user to content recommendation, it is determined that with
Satisfaction of the family to the content recommendation；
3rd satisfaction determining unit, for according to support feedback data of the user to content recommendation and concern time, really
Determine satisfaction of the user to the content recommendation.
Further, described device also includes：
Pushing module, for prompting the interest prediction result to applications client, and user is obtained to the interest
The click data of prediction result, after determining user interest, according to the user interest of determination, correct the recommendation pushed to user
Content.
The advantageous effects of technical scheme that the embodiment of the present invention proposes are：By obtain user input in recommendation
The expression data or behavioral data of appearance, interest prediction result is determined, then to applications client reminding interest prediction result, and obtain
User is to the click data of interest prediction result, to determine user interest.The embodiment of the present invention can allow for user actively to pass through
Natural language carries out interest expression, finds user interest accurately and in time, lifts Consumer's Experience.
Brief description of the drawings
In order to illustrate more clearly of the present invention, one will be done to the required accompanying drawing used in the present invention below and be simply situated between
Continue, it should be apparent that, drawings in the following description are some embodiments of the present invention, are come for those of ordinary skill in the art
Say, without having to pay creative labor, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is the flow chart for the user interest discovery method that the specific embodiment of the invention one provides；
Fig. 2 is the flow chart for the user interest discovery method that the specific embodiment of the invention two provides；
Fig. 3 is the flow chart for the user interest discovery method that the specific embodiment of the invention three provides；
Fig. 4 is the flow chart for the user interest discovery method that the specific embodiment of the invention four provides；
Fig. 5 is the flow chart for the user interest discovery method that the specific embodiment of the invention five provides；
Fig. 6 is the flow chart for the user interest discovery method that the specific embodiment of the invention five provides；
Fig. 7 is the flow chart for the user interest discovery method that the specific embodiment of the invention five provides；
Fig. 8 is the structured flowchart that the user interest that the specific embodiment of the invention six provides finds device.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to the embodiment of the present invention
In technical scheme be described in further detail, it is clear that described embodiment is part of the embodiment of the present invention, rather than entirely
The embodiment in portion.It is understood that specific embodiment described herein is only used for explaining the present invention, rather than to the present invention's
Limit, based on the embodiment in the present invention, those of ordinary skill in the art are obtained under the premise of creative work is not made
Every other embodiment, belong to the scope of protection of the invention.It also should be noted that for the ease of description, accompanying drawing
In illustrate only part related to the present invention rather than full content.
Embodiment one
Fig. 1 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
101st, the expression data to content recommendation of user's input are obtained.
102nd, according to the expression data, it is determined that object of interest corresponding with the expression data, including：
The word that the expression data are included is matched in object repository；
Using the word that the match is successful as object of interest corresponding to the expression data.
103rd, according to the expression data, it is determined that user's attitude corresponding with the expression data, including：
The word that the expression data are included is matched in default emotional attitude template；
According to matching result, it is determined that user's attitude corresponding with the expression data.
104th, according to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.
105th, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
In the present embodiment, expressing data, concretely user inputs institute's table using natural language by voice or word
The hope increase reached, the interest expression data for reducing or shielding some content recommendations.Such as：" I is wanted to see more on 99 formulas
The news of tank ", " go out again hammer mobile phone news I just unload！", " news of millet is too many, and you have collected money" etc.
Deng user can use text input mode or phonetic entry pattern, and text input mode is common input pattern, herein not
It is described in detail；And phonetic entry pattern, the voice of user can be converted into text message by sound identification module first, this step
Rapid completion has the solution of many maturations, such as existing speech recognition open platform.As long as recommendation service uses these
Freeware development kit (the SDK of speech-recognition services business：Software Development Kit) just can be freely sharp
With the speech-recognition services of these platforms, and this technology is highly developed in putting into practice, and recognition accuracy is up to more than 95%.
The intention that user is gone out with natural language expressing is identified, its Major Difficulties is according to corresponding to determining expression data
Object of interest and user's attitude, i.e. step 102 and step 103.It is two tuples by the intention assessment problem reduction of user
Identification：
<Object,Attitude>
Wherein Object represents the object of interest that user is directed to, and object of interest in the above example is " 99 formulas respectively
Tank ", " hammer mobile phone ", " millet (millet Science and Technology Ltd.) ".Attitude represents user's attitude, can use limited collection
Close to represent, such as：{ ' shielding ', ' reduce and recommend ', ' increase is recommended ' }.
In actual applications, Object determination needs to rely on an Object knowledge base, i.e. object repository.Object is known
Knowing storehouse needs the significant Object of a large amount of manual confirmations, in knowledge base comprising the alias (ref) corresponding to object of interest,
The title (name) of type (type) and object of interest, wherein type can be classification (category) or entity
(entity：Name, mechanism name, make the specific things such as the name of an article).Such as：
<ref:[science and technology news, scientific and technical article, science and technology], type：Category, name:Science and technology>
<ref:[millet, millet science and technology], type：Entity, name:Millet Science and Technology Ltd.>
<ref:[Fan Bingbing], type：Entity, name:Fan Bingbing>
There is object repository, it is possible to according to the Object in alias mapping one section of word of identification.
Attitude determination mainly according to the rule of some Manual definitions, such as：
Carry out some XXX-- ' less and reduce recommendation '
Recommend point XXX-- ' increases to recommend ' more
Should not XXX-- ' shielding '
It should be noted that in the case of having one kind, some word in one section of word is possible to map multiple Object, such as
' millet ' is possible to refer to millet Science and Technology Ltd., it is also possible to refers to edible millet, can prompt user in this case, allow
User confirms that to be expressed is any meaning, or judges which Object and user more phase by the usage record of user
Close.For example if user's expression is negative attitude, it is more likely that user thinks to appear in the Object that user is browsed in history
The object of attitude is expressed, the Object thus expressed to confirm user really to want.
Also in the case of one kind, the Object in the intention of user is not a specific concept, but relatively more abstract
Concept.Such as ' it is interesting to come some ', ' not go out again vulgar ' is such to express.Here ' interesting ' and ' vulgar ' all belongs to
In comparing abstract concept, some specific Object can not be corresponded to.In such a case, it is possible to it is special to pre-define some
Different label, such as ' cross-talk ', ' vulgar content ' are used as Object.These labels can be marked by operation personnel to content
Note, can also use machine learning algorithm automatic identification.So, demand relatively more abstract in this expression can be also identified,
And then determine Object.
Next, according to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.To
Applications client prompts the interest prediction result, and user selects to click on one of interest prediction result, obtains user to institute
The click data of interest prediction result is stated, to determine user interest.
Such as：Above citing " go out again hammer mobile phone news I just unload！", it may be determined that Object is " hammer hand
Machine ", Attitude are " shielding ", then can be supplied to applications client reminding interest prediction result for " shielding hammer mobile phone news "
User clicks on, if the prediction result does not meet user's true intention, user can be neglected；Or " shielding hammer mobile phone is new for prompting
Hear", there is "Yes" and "No" to supply user to select, user selects to click on one of interest prediction result, obtains user to institute
The click data of interest prediction result is stated, to determine user interest.
For another example：Above illustrate " I wants to see the more news on 99 formula tanks ", it may be determined that Object is " 99 formulas are smooth
Gram ", Attitude is " increase is recommended ", then " can recommend the research and development of 99 formula tanks to applications client reminding interest prediction result
Multiple choosings such as progress news ", " recommend 99 formula tanks come into operation situation news " and " recommendation western medium is on 99 formula tank news "
Item selects for user, and user selects to click on one of interest prediction result, point of the acquisition user to the interest prediction result
Data are hit, to determine user interest.
Present embodiments provide a kind of user interest discovery method, this method is by obtaining that user inputs to content recommendation
Expression data, determine interest prediction result, then to applications client reminding interest prediction result, and it is pre- to interest to obtain user
The click data of result is surveyed, to determine user interest.User can actively express requirement to content recommendation rather than passive
Waiting system understands the interest of oneself by very long data mining process, and user can pass through natural language expressing oneself
Demand, running cost are very low.The present embodiment can find user interest accurately and in time, lift Consumer's Experience.
Embodiment two
Fig. 2 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
201st, the expression data to content recommendation of user's input are obtained.
202nd, according to the expression data, it is determined that object of interest corresponding with the expression data, including：
The word that the expression data are included is matched in object repository；
Using the word that the match is successful as object of interest corresponding to the expression data.
203rd, according to the expression data, it is determined that user's attitude corresponding with the expression data, including：
The word that the expression data are included is matched in default emotional attitude template；
According to matching result, it is determined that user's attitude corresponding with the expression data.
204th, according to the text distance between object of interest word corresponding with user's attitude, filter out text away from
From the object of interest more than setting value.
205th, according to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.
206th, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
In the present embodiment, expressing data, concretely user inputs institute's table using natural language by voice or word
The hope increase reached, the interest expression data for reducing or shielding some content recommendations.User can use text input mode or
Person's phonetic entry pattern.
The intention that user is gone out with natural language expressing is identified, its Major Difficulties is according to corresponding to determining expression data
Object of interest and user's attitude, i.e. step 202 and step 203.It is two tuples by the intention assessment problem reduction of user
Identification：
<Object,Attitude>
Wherein Object represents the object of interest that user is directed to, and Attitude represents user's attitude, can use limited collection
Close to represent, such as：{ ' shielding ', ' reduce and recommend ', ' increase is recommended ' }.
In actual applications, Object determination needs to rely on an Object knowledge base, i.e. object repository.Object is known
Knowing storehouse needs the significant Object of a large amount of manual confirmations, in knowledge base comprising the alias (ref) corresponding to object of interest,
The title (name) of type (type) and object of interest, wherein type can be classification (category) or entity
(entity：Name, mechanism name, make the specific things such as the name of an article).There is object repository, it is possible to according to alias mapping identification one
Object in section words.
Attitude determination mainly according to the rule of some Manual definitions, such as：
Carry out some XXX-- ' less and reduce recommendation '
Recommend point XXX-- ' increases to recommend ' more
Should not XXX-- ' shielding '
It should be noted that in the case of having one kind, multiple Object, such as " millet news are can recognize that in one section of word
It is too many, it is advertisement", " millet news " and " advertisement " in the words can find corresponding Object.Such case
Under, it can judge with reference to Attitude identification, i.e. step 204, if Object and Attitude text
Distance is more than setting value, can filter out the Object, and the Object thus expressed to confirm user really to want, setting value can be with
It is that user interest finds that device is pre-set or voluntarily set by user, if setting value is " 2 ", one
When Object and Attitude text distance is more than 2, the Object can be filtered out, " too many " can determine in upper example
For Attitude, the text distance of " millet news " between " too many " is 0, and the text distance between " advertisement " and " too many "
For 3, then " advertisement " this Object can be filled into, to confirm that user really thinks " the millet news " of expression.
Next, according to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.To
Applications client prompts the interest prediction result, and user selects to click on one of interest prediction result, obtains user to institute
The click data of interest prediction result is stated, to determine user interest.
A kind of user interest discovery method is present embodiments provided, this method adds root on the basis of embodiment one
According to the text distance between object of interest word corresponding with user's attitude, the interest pair that text distance is more than setting value is filtered out
As to improve the accuracy rate of object of interest determination.The present embodiment can find user interest accurately and in time, lift user's body
Test.
Embodiment three
Fig. 3 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
301st, the behavioral data to content recommendation of user's input is obtained；
302nd, according to the behavioral data, satisfaction of the user to the content recommendation is determined.Wherein it is determined that user is to institute
A variety of implementations can be had by stating the satisfaction of content recommendation, including at least one of following：
Refreshing frequency according to user to content recommendation, determine satisfaction of the user to the content recommendation；
Click data and stay time according to user to content recommendation, determine satisfaction of the user to the content recommendation
Degree；
According to support feedback data of the user to content recommendation and concern time, determine that user expires to the content recommendation
Meaning degree.
303rd, judge whether the satisfaction is less than setting threshold value, if the satisfaction is less than setting threshold value,
Triggering performs the operation for the expression data to content recommendation for obtaining user's input, that is, performs step 304.
304th, the expression data to content recommendation of user's input are obtained.
305th, according to the expression data, it is determined that object of interest corresponding with the expression data, including：
The word that the expression data are included is matched in object repository；
Using the word that the match is successful as object of interest corresponding to the expression data.
306th, according to the expression data, it is determined that user's attitude corresponding with the expression data, including：
The word that the expression data are included is matched in default emotional attitude template；
According to matching result, it is determined that user's attitude corresponding with the expression data.
307th, according to the text distance between object of interest word corresponding with user's attitude, filter out text away from
From the object of interest more than setting value.
308th, according to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.
309th, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
In the present embodiment, behavioral data concretely refreshing frequency of the user to content recommendation, the point to content recommendation
Hit corresponding to the behavior in use such as data and stay time, the support feedback data to content recommendation and concern time
Data, determine satisfaction of the user to the content recommendation according to these behavioral datas, set satisfaction threshold value, if institute is really
Fixed satisfaction is higher than threshold value, can continue to recommend, if satisfaction is less than threshold value, triggering performs pair for obtaining user's input
The operation of the expression data of content recommendation.Follow-up operation is identical with embodiment two, and no further details to be given herein.
As further explanation, the refreshing frequency according to user to content recommendation in the present embodiment, determine user to institute
Stating the mode of the satisfaction of content recommendation can be：When user is unsatisfied with to content recommendation, user read content recommendation when
Between will be short, and the height of the length of reading time and refreshing frequency is inversely proportional substantially, when full corresponding to the high phase of refreshing frequency
Meaning degree is low, and satisfaction corresponding to the low phase of refreshing frequency is high, it is assumed that refreshing frequency be 5 times per hour corresponding satisfaction be door
Limit value, when the refreshing frequency of user is hourly for 6 times when, will trigger perform obtain user's input to content recommendation
Express the operation of data；
Click data and stay time according to user to content recommendation, determine satisfaction of the user to the content recommendation
Mode can be：Using grading scheme, recommend article when user clicks on one, illustrate that user is interested in this article, Ke Yiwei
The content is scored, such as 1 point, and when user reads this article, reading time is scored, and is such as often read 30 seconds and is remembered 1 point, reads 2 points
Clock is 4 points, of course, it is possible to be adjusted according to article content length to obatained score, is not described in detail herein, it is assumed that tired
The fraction of 10 recommendation articles of meter is the score value of satisfaction, and threshold value is 20 points, full when user, which have ignored more, recommends article
Meaning degree score value will be very low, when satisfaction score value is less than 20 timesharing, will trigger perform obtain user's input to content recommendation
Expression data operation；
According to support feedback data of the user to content recommendation and concern time, determine that user expires to the content recommendation
The mode of meaning degree can be：Customer satisfaction survey option can be provided with every recommendation article and sets corresponding score value, such as
" very satisfied " is 5 points, and " feeling quite pleased " is 4 points, and " general " is 3 points, and " dissatisfied " is 2 points, and " very dissatisfied " is 1 point, is used
Family draws support feedback data of the user to content recommendation, institute by selecting one of option to recommend essay grade for the piece
It can be concern time of the user to the content recommendation in a certain field to state the concern time, such as amusement, physical culture and economy, concern
Time grows the satisfaction score that can then increase the field content recommendation, and finally in summary two scores draw final satisfaction
Score value, when satisfaction score value is less than threshold value, the expression data to content recommendation for performing and obtaining user's input will be triggered
Operation.The citing of embodiment is these are only, is not limited to the embodiment of the above in actual applications.
A kind of user interest discovery method is present embodiments provided, this method is added and obtained on the basis of embodiment two
The behavioral data to content recommendation of family input is taken, and according to the behavioral data, determines user to the content recommendation
The step of satisfaction, if satisfaction is higher, the step of user actively carries out data representation can be reduced, further lifts user's body
Test.
Example IV
Fig. 4 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
401st, the behavioral data to content recommendation of user's input is obtained；
402nd, according to the behavioral data, satisfaction of the user to the content recommendation is determined.
403rd, judge whether the satisfaction is less than setting threshold value, if the satisfaction is less than setting threshold value,
Triggering performs the operation for the expression data to content recommendation for obtaining user's input.
404th, the expression data to content recommendation of user's input are obtained.
405th, according to the expression data, it is determined that object of interest corresponding with the expression data, including：
The word that the expression data are included is matched in object repository；
Using the word that the match is successful as object of interest corresponding to the expression data.
406th, according to the expression data, it is determined that user's attitude corresponding with the expression data, including：
The word that the expression data are included is matched in default emotional attitude template；
According to matching result, it is determined that user's attitude corresponding with the expression data.
407th, according to the text distance between object of interest word corresponding with user's attitude, filter out text away from
From the object of interest more than setting value.
408th, according to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.
409th, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
410th, according to the user interest of determination, the content recommendation pushed to user is corrected.
Modification rule can be formulated for the amendment, and modification rule can have a variety of, and only citing illustrates herein.Such as
It can be that each article sets initial score, the fraction of article is adjusted correspondingly according to user interest, fraction is high
Article is easier recommended, and the low article of fraction is more difficult to be recommended, and the article that fraction is zero can be shielded.Such as：For
User interest<' millet ', ' reduce and recommend '>, can trigger and " the article fraction reduction by 50% " of millet, " keyword be included in title
In comprising millet article fraction reduce by 20% " two modification rule.After modification rule comes into force, the article fraction of " millet " is hit
Can be penalized, thus recommended come out can be more difficult to.The number that the entry-into-force time of modification rule can express according to user determines.Such as
Fruit is that user expresses for the first time, and the entry-into-force time can be three days, and second can be one week, can be one month for the third time；And
For<' Huawei ', ' increase is recommended '>, the modification rule of " being doubled in title comprising the article fraction of Huawei " can be triggered,
Certainly, modification rule can also be other forms, such as：" if the article quantity of Huawei is included in recommended candidate article in title
Recommended candidate is added from 5 such articles of database search " less than 5.Furthermore, it is possible to actively expressed according to user
Frequency adjusts modification rule, the user is quickly seen the improvement of recommendation results.
A kind of user interest discovery method is present embodiments provided, this method adds root on the basis of embodiment three
According to the user interest of determination, the step of correcting the content recommendation pushed to user, the step can be such that the expression of user gives birth to immediately
Effect, user can further lift Consumer's Experience with regard to that can see the improvement of recommendation results when recommending after expression next time.
Embodiment five
Fig. 5 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
501st, the behavioral data to content recommendation of user's input is obtained.
502nd, according to the behavioral data, interest prediction result is determined.
503rd, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
In the present embodiment, behavioral data concretely refreshing frequency of the user to content recommendation, the point to content recommendation
Hit corresponding to the behavior in use such as data and stay time, the support feedback data to content recommendation and concern time
Data, as further explanation, when user is unsatisfied with to content recommendation, the time that user reads content recommendation will be short,
And the length of reading time and the height of refreshing frequency are inversely proportional substantially, illustrate user to content recommendation not when refreshing frequency is high
It is interested, and illustrate that user is interested in content recommendation when refreshing frequency is low, selected with this determination interest prediction result for user；
Recommend article when user clicks on one, illustrate that user is interested in this article, it is right when user, which reads this article, spends the time long
This article is very interested, is selected with this determination interest prediction result for user；It can recommend to be provided with use in article at every in addition
Family satisfaction investigation option, as " very satisfied ", " feeling quite pleased ", " ", " dissatisfied " and " very dissatisfied ", user passes through
One of option is selected, draws support feedback data of the user to content recommendation, the concern time can be user to certain
The concern time of the content recommendation in one field, such as amusement, physical culture and economy, concern time length then illustrate user to the field
Content recommendation is interested, finally determines that interest prediction result selects for user.The citing of embodiment is these are only, in reality
The embodiment of the above is not limited in the application of border.
Such as：If user have ignored the more recommendation articles on millet, then user is likely to that it is emerging not feel millet
Interest.If user is more to the article number of clicks on electric automobile, then user is likely to Tesla (CS) Koncern, Podebradska 186, Praha 9, Czechoslovakia founder
The news of Maas gram is also interested.Traditionally, these information can be used directly to reduce or more recommend certain class news.But
It is that, because the behavior of user is complicated, this prediction is likely to be mistake.Blindness recommends article very according to these predictions
The satisfaction of user may not be lifted.So interest prediction result can be determined according to the behavioral data, to application visitor
Family end reminding interest prediction result, and click data of the user to the interest prediction result is obtained, to determine user interest.Than
Such as, can to the following interest prediction result of Client-Prompt,
Whether you think：
Reduce the recommendation on millet
Recommend the news of Maas gram
Shield the news for net of cutting firewood
If there is a wish for meeting user in above-mentioned prediction result, user can click on corresponding interest prediction result
Determine user interest.
A kind of user interest discovery method is present embodiments provided, this method determines that interest is pre- according to user behavior data
Result is surveyed, prompts the interest prediction result to applications client, and obtain hits of the user to the interest prediction result
According to determine user interest.The present embodiment can predict the possible recommended requirements of user according to user behavior data, once prediction
Success, user only need to confirm, can save voice or word input step, can further lift Consumer's Experience.
Embodiment six
Fig. 6 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
601st, the behavioral data to content recommendation of user's input is obtained.
602nd, according to the behavioral data, satisfaction of the user to the content recommendation is determined.
603rd, judge whether the satisfaction is less than setting threshold value, if so, then triggering is performed according to the behavioral data,
The operation of interest prediction result is determined, that is, performs step 604.
604th, according to the behavioral data, interest prediction result is determined.
605th, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
In the present embodiment, behavioral data concretely refreshing frequency of the user to content recommendation, the point to content recommendation
Hit corresponding to the behavior in use such as data and stay time, the support feedback data to content recommendation and concern time
Data, determine satisfaction of the user to the content recommendation according to these behavioral datas, set satisfaction threshold value, if institute is really
Fixed satisfaction is higher than threshold value, can continue to recommend, if satisfaction is less than threshold value, triggering is performed according to the behavior number
According to determining interest prediction result.
A kind of user interest discovery method is present embodiments provided, this method determines user to recommending according to behavioral data
The satisfaction of content, if satisfaction is less than setting threshold value, triggering is performed according to behavioral data, determines interest prediction knot
Fruit, the interest prediction result is prompted to applications client, and obtain click data of the user to the interest prediction result, with
Determine user interest.The number for the step of user actively carries out data representation can be reduced, greatly reduces use cost, simultaneously
Consumer's Experience can further be lifted.
Embodiment seven
Fig. 7 is a kind of flow chart for user interest discovery method that the present embodiment provides, and this method comprises the following steps：
701st, the behavioral data to content recommendation of user's input is obtained.
702nd, according to the behavioral data, interest prediction result is determined.
703rd, the interest prediction result is prompted to applications client, and obtains point of the user to the interest prediction result
Data are hit, to determine user interest.
704th, according to the user interest of determination, the content recommendation pushed to user is corrected.
The present embodiment adds the user interest according to determination on the basis of embodiment five, corrects and is pushed away to what user pushed
The step for recommending content, it can further lift Consumer's Experience.
Embodiment eight
Based on the various embodiments described above, the present embodiment provides a kind of user interest and finds device, as shown in figure 8, being this
The structured flowchart of embodiment device, the device may include：
User data acquisition module 801, for obtaining the expression data or behavioral data to content recommendation of user's input.
Interest prediction result determining module 802, for according to the expression data or the behavioral data, determining that interest is pre-
Result is surveyed, including：Object of interest determining unit, for according to the expression data, it is determined that corresponding emerging with the expression data
Interesting object；User's attitude determining unit, for according to the expression data, it is determined that User space corresponding with the expression data
Degree；Interest prediction result determining unit, for according to object of interest corresponding with the expression data and user's attitude, determining emerging
Interesting prediction result.Wherein, object of interest determining unit specifically includes：First coupling subelement, in object repository
The word included with the expression data；Object of interest determination subelement, for using the word that the match is successful as the expression number
According to corresponding object of interest；Object of interest filters subelement, for the word that the match is successful is corresponding as the expression data
Object of interest after, according to the text distance between object of interest word corresponding with user's attitude, filter out text
This distance is more than the object of interest of setting value.Wherein, user's attitude determining unit includes：Second coupling subelement, for pre-
If the word that the expression data are included is matched in emotional attitude template；User's attitude determination subelement, for being tied according to matching
Fruit, it is determined that user's attitude corresponding with the expression data.
Interest determination module 803, for prompting the interest prediction result to applications client, and user is obtained to described
The click data of interest prediction result, to determine user interest.
First satisfaction determining module 804, for obtain user input the expression data to content recommendation before, root
According to the behavioral data, satisfaction of the user to the content recommendation is determined, including it is at least one of following：First satisfaction determines
Unit, for the refreshing frequency according to user to content recommendation, determine satisfaction of the user to the content recommendation；Second satisfaction
Determining unit is spent, for the click data and stay time according to user to content recommendation, determines user to the content recommendation
Satisfaction；3rd satisfaction determining unit, for according to support feedback data of the user to content recommendation and concern time, really
Determine satisfaction of the user to the content recommendation.
Data acquisition trigger module 805 is expressed, if being less than setting threshold value for the satisfaction, triggering execution obtains
Take the operation of the expression data to content recommendation of family input.
Second satisfaction determining module 806, for according to the behavioral data, before determining interest prediction result, root
According to the behavioral data, satisfaction of the user to the content recommendation is determined, including it is at least one of following：First satisfaction determines
Unit, for the refreshing frequency according to user to content recommendation, determine satisfaction of the user to the content recommendation；Second satisfaction
Determining unit is spent, for the click data and stay time according to user to content recommendation, determines user to the content recommendation
Satisfaction；3rd satisfaction determining unit, for according to support feedback data of the user to content recommendation and concern time, really
Determine satisfaction of the user to the content recommendation.
Interest prediction result determines trigger module 807, if being less than setting threshold value for the satisfaction, triggering is held
Row determines interest prediction result according to the behavioral data.
Pushing module 808, for prompting the interest prediction result to applications client, and user is obtained to described emerging
The click data of interesting prediction result, after determining user interest, according to the user interest of determination, correct and pushed away to what user pushed
Recommend content.
A kind of user interest provided in an embodiment of the present invention finds that device can perform what any embodiment of the present invention was provided
User interest discovery method, possess the corresponding functional module of execution method and beneficial effect.
Finally it should be noted that：Various embodiments above is merely to illustrate technical scheme, rather than it is limited
System；Preferred embodiment in embodiment, is not limited, and to those skilled in the art, the present invention can be with
There are various changes and change.All any modification, equivalent substitution and improvements made within spirit and principles of the present invention etc.,
It should be included within protection scope of the present invention.
Claims (16)
- A kind of 1. user interest discovery method, it is characterised in that including：The expression data or behavioral data to content recommendation of user's input are obtained, wherein the expression data are that user uses certainly The interest expression data of increase, reduction or shielding content recommendation expressed by right language；According to the expression data or the behavioral data, interest prediction result is determined；The interest prediction result is prompted to applications client, and obtains click data of the user to the interest prediction result, To determine user interest；Wherein, according to the expression data, interest prediction result is determined, including：According to the expression data, it is determined that object of interest corresponding with the expression data and user's attitude；According to object of interest corresponding with the expression data and user's attitude, interest prediction result is determined.
- 2. according to the method for claim 1, it is characterised in that obtaining the expression data to content recommendation of user's input Before, in addition to：Obtain the behavioral data to content recommendation of user's input；According to the behavioral data, satisfaction of the user to the content recommendation is determined；If the satisfaction is less than setting threshold value, triggering performs the expression data to content recommendation for obtaining user's input Operation.
- 3. according to the method for claim 1, it is characterised in that according to the expression data, it is determined that with the expression data Corresponding object of interest, including：The word that the expression data are included is matched in object repository；Using the word that the match is successful as object of interest corresponding to the expression data.
- 4. according to the method for claim 3, it is characterised in that the word that the match is successful is corresponding as the expression data Object of interest after, in addition to：According to the text distance between object of interest word corresponding with user's attitude, filter out text distance and be more than and set The object of interest of definite value.
- 5. according to the method for claim 1, it is characterised in that according to the expression data, it is determined that with the expression data Corresponding user's attitude, including：The word that the expression data are included is matched in default emotional attitude template；According to matching result, it is determined that user's attitude corresponding with the expression data.
- 6. according to the method for claim 1, it is characterised in that according to the behavioral data, determine interest prediction result Before, in addition to：According to the behavioral data, satisfaction of the user to the content recommendation is determined；If the satisfaction is less than setting threshold value, triggering is performed according to the behavioral data, determines interest prediction result Operation.
- 7. the method according to claim 2 or 6, it is characterised in that according to the behavioral data, determine that user pushes away to described The satisfaction of content is recommended, including it is at least one of following：Refreshing frequency according to user to content recommendation, determine satisfaction of the user to the content recommendation；Click data and stay time according to user to content recommendation, determine satisfaction of the user to the content recommendation；According to support feedback data of the user to content recommendation and concern time, satisfaction of the user to the content recommendation is determined Degree.
- 8. according to any described methods of claim 1-6, it is characterised in that prompting the interest to predict to applications client As a result, and click data of the user to the interest prediction result is obtained, after determining user interest, in addition to：According to the user interest of determination, the content recommendation pushed to user is corrected.
- 9. a kind of user interest finds device, it is characterised in that including：User data acquisition module, for obtaining the expression data or behavioral data to content recommendation of user's input, wherein institute It is that user uses the increase expressed by natural language, reduction or the interest expression data for shielding content recommendation to state expression data；Interest prediction result determining module, for according to the expression data or the behavioral data, determining interest prediction result；Interest determination module, for prompting the interest prediction result to applications client, and it is pre- to the interest to obtain user The click data of result is surveyed, to determine user interest；Wherein, the interest prediction result determining module includes：Object of interest determining unit, for according to the expression data, it is determined that object of interest corresponding with the expression data；User's attitude determining unit, for according to the expression data, it is determined that user's attitude corresponding with the expression data；Interest prediction result determining unit, for basis object of interest corresponding with the expression data and user's attitude, it is determined that Interest prediction result.
- 10. device according to claim 9, it is characterised in that described device also includes：First satisfaction determining module, for obtain user input the expression data to content recommendation before, according to described Behavioral data, determine satisfaction of the user to the content recommendation；Data acquisition trigger module is expressed, for when the satisfaction is less than setting threshold value, it is defeated that triggering execution obtains user The operation of the expression data to content recommendation entered.
- 11. device according to claim 9, it is characterised in that object of interest determining unit includes：First coupling subelement, the word included for matching the expression data in object repository；Object of interest determination subelement, for using the word that the match is successful as object of interest corresponding to the expression data.
- 12. device according to claim 11, it is characterised in that object of interest determining unit also includes：Object of interest filters subelement, for using the word that the match is successful as object of interest corresponding to the expression data it Afterwards, according to the text distance between object of interest word corresponding with user's attitude, filter out text distance and be more than and set The object of interest of definite value.
- 13. device according to claim 9, it is characterised in that user's attitude determining unit includes：Second coupling subelement, the word included for matching the expression data in default emotional attitude template；User's attitude determination subelement, for according to matching result, it is determined that user's attitude corresponding with the expression data.
- 14. device according to claim 9, it is characterised in that described device also includes：Second satisfaction determining module, for according to the behavioral data, before determining interest prediction result, according to the row For data, satisfaction of the user to the content recommendation is determined；Interest prediction result determines trigger module, for when the satisfaction is less than setting threshold value, triggering to be performed according to institute Behavioral data is stated, determines the operation of interest prediction result.
- 15. device according to claim 10, it is characterised in that the first satisfaction determining module includes following at least one ：First satisfaction determining unit, for the refreshing frequency according to user to content recommendation, determine user in the recommendation The satisfaction of appearance；Second satisfaction determining unit, for the click data and stay time according to user to content recommendation, determine user couple The satisfaction of the content recommendation；3rd satisfaction determining unit, for according to support feedback data of the user to content recommendation and concern time, it is determined that with Satisfaction of the family to the content recommendation.
- 16. according to any described devices of claim 9-15, it is characterised in that described device also includes：Pushing module, for prompting the interest prediction result to applications client, and obtain user and the interest is predicted As a result click data, after determining user interest, according to the user interest of determination, correct in the recommendation pushed to user Hold.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410613040.3A CN104361063B (en) | 2014-11-04 | 2014-11-04 | user interest discovery method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410613040.3A CN104361063B (en) | 2014-11-04 | 2014-11-04 | user interest discovery method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104361063A CN104361063A (en) | 2015-02-18 |
| CN104361063B true CN104361063B (en) | 2018-03-16 |
Family
ID=52528324
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410613040.3A Active CN104361063B (en) | 2014-11-04 | 2014-11-04 | user interest discovery method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104361063B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105303398B (en) * | 2015-09-29 | 2020-03-27 | 努比亚技术有限公司 | Information display method and system |
| CN105302416A (en) * | 2015-10-29 | 2016-02-03 | 崔东珠 | Cartoon exhibition system |
| CN105512224A (en) * | 2015-11-30 | 2016-04-20 | 清华大学 | Search engine user satisfaction automatic assessment method based on cursor position sequence |
| KR102362868B1 (en) * | 2015-12-23 | 2022-02-15 | 삼성전자주식회사 | A method for providing contents to a user based on preference of the user and an electronic device therefor |
| CN107870912A (en) * | 2016-09-22 | 2018-04-03 | 广州市动景计算机科技有限公司 | Article quality score method, equipment, client, server and programmable device |
| CN106850762B (en) * | 2017-01-03 | 2021-12-14 | 腾讯科技（深圳）有限公司 | Message pushing method, server and message pushing system |
| CN108305091B (en) * | 2017-01-13 | 2022-04-26 | 北京京东尚科信息技术有限公司 | Electronic equipment, user interest perception degree extraction method and device |
| CN108733666B (en) * | 2017-04-13 | 2022-03-08 | 腾讯科技（深圳）有限公司 | Server information pushing method, terminal information sending method, device and system |
| US11270246B2 (en) | 2017-06-05 | 2022-03-08 | Accenture Global Solutions Limited | Real-time intelligent and dynamic delivery scheduling |
| CN107273489B (en) * | 2017-06-14 | 2019-08-30 | 掌阅科技股份有限公司 | Content delivery method, electronic equipment and computer storage medium |
| CN107766467B (en) * | 2017-09-29 | 2020-04-17 | 北京金山安全软件有限公司 | Information detection method and device, electronic equipment and storage medium |
| CN107886357A (en) * | 2017-11-06 | 2018-04-06 | 北京希格斯科技发展有限公司 | The method and system of content value is judged based on user behavior data |
| CN107911491B (en) * | 2017-12-27 | 2019-09-27 | Oppo广东移动通信有限公司 | Information recommendation method, device and storage medium, server and mobile terminal |
| CN110516159B (en) * | 2019-08-30 | 2022-12-20 | 北京字节跳动网络技术有限公司 | Information recommendation method and device, electronic equipment and storage medium |
| CN111444438B (en) * | 2020-03-24 | 2023-09-01 | 北京百度网讯科技有限公司 | Method, device, equipment and storage medium for determining quasi-recall rate of recall strategy |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663001A (en) * | 2012-03-15 | 2012-09-12 | 华南理工大学 | Automatic blog writer interest and character identifying method based on support vector machine |
| CN103324742A (en) * | 2013-06-28 | 2013-09-25 | 百度在线网络技术（北京）有限公司 | Method and equipment for recommending keywords |
| CN103780625A (en) * | 2014-01-26 | 2014-05-07 | 北京搜狗科技发展有限公司 | Method and device for discovering interest of users |
| CN103888466A (en) * | 2014-03-28 | 2014-06-25 | 北京搜狗科技发展有限公司 | User interest discovering method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8438170B2 (en) * | 2006-03-29 | 2013-05-07 | Yahoo! Inc. | Behavioral targeting system that generates user profiles for target objectives |
-
2014
- 2014-11-04 CN CN201410613040.3A patent/CN104361063B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663001A (en) * | 2012-03-15 | 2012-09-12 | 华南理工大学 | Automatic blog writer interest and character identifying method based on support vector machine |
| CN103324742A (en) * | 2013-06-28 | 2013-09-25 | 百度在线网络技术（北京）有限公司 | Method and equipment for recommending keywords |
| CN103780625A (en) * | 2014-01-26 | 2014-05-07 | 北京搜狗科技发展有限公司 | Method and device for discovering interest of users |
| CN103888466A (en) * | 2014-03-28 | 2014-06-25 | 北京搜狗科技发展有限公司 | User interest discovering method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104361063A (en) | 2015-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104361063B (en) | user interest discovery method and device | |
| CN110609902B (en) | Text processing method and device based on fusion knowledge graph | |
| US10235470B2 (en) | User retrieval enhancement | |
| JP6629246B2 (en) | Learning and Using Context-Aware Content Acquisition Rules for Query Disambiguation | |
| US20120089641A1 (en) | Searching travel records | |
| US8275666B2 (en) | User supplied and refined tags | |
| CN109582799B (en) | Method and device for determining knowledge sample data set and electronic equipment | |
| CN106970991B (en) | Similar application identification method and device, application search recommendation method and server | |
| US11392645B2 (en) | Computerized domain expert | |
| US10585927B1 (en) | Determining a set of steps responsive to a how-to query | |
| US11416565B2 (en) | Techniques to leverage machine learning for search engine optimization | |
| CN107861753B (en) | APP generation index, retrieval method and system and readable storage medium | |
| CN109255012B (en) | Method and device for machine reading understanding and candidate data set size reduction | |
| CN106462613A (en) | Ranking suggestions based on user attributes | |
| CN111339284A (en) | Product intelligent matching method, device, equipment and readable storage medium | |
| US11138240B2 (en) | Personalized automatic tagging | |
| CN111861619A (en) | Recommendation method and system for shared vehicles | |
| CN103235777B (en) | A kind of internet public feelings acquisition methods of user demand guidance | |
| US11488223B1 (en) | Modification of user interface based on dynamically-ranked product attributes | |
| CN108470289B (en) | Virtual article issuing method and equipment based on E-commerce shopping platform | |
| CN110235121A (en) | For the enhanced system and method investigated online | |
| CN116186413A (en) | Method and system for multi-modal feature recommendation | |
| CN105975508A (en) | Personalized meta-search engine searched result merging and sorting method | |
| EP3731108A1 (en) | Search system, search method, and program | |
| CN113486248A (en) | Recommendation method and device for functional module in application program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |