CN104361063A - User interest discovering method and device - Google Patents
User interest discovering method and device Download PDFInfo
- Publication number
- CN104361063A CN104361063A CN201410613040.3A CN201410613040A CN104361063A CN 104361063 A CN104361063 A CN 104361063A CN 201410613040 A CN201410613040 A CN 201410613040A CN 104361063 A CN104361063 A CN 104361063A
- Authority
- CN
- China
- Prior art keywords
- user
- interest
- content recommendation
- data
- expression data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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 invention provides a user interest discovering method and a device. The user interest discovering method comprises the following steps of obtaining expressing data or behavior data input by a user to recommending content, determining an interest predicting result according to the expressing data or the behavior data, prompting the interest predicting result to an application client and obtaining clicking data of the user to the interest predicting result to determine the interest of the user. The user interest discovering method and the device can allow the user to initiatively express the interest through a natural language and accurately discover the interest of the user in time, and the user experience is improved.
Description
Technical field
The embodiment of the present invention relates to areas of information technology, particularly relates to a kind of user interest discover method and device.
Background technology
Recommendation of personalized information technology can issue to user the information meeting user interest, and therefore, this technology is more and more applied gradually in network access.In recommendation of personalized information technology, need discovery user interest accurately and timely.
Existing user interest discovery technique, is generally obtain user to the click of content recommendation, shares, positive feedback behaviors such as collection and/or ignore, and the negative feedback behavior such as to step on, and analyze user interest from feedback behavior.
There is following defect in above-mentioned user interest discovery technique: on the one hand, the information of the feedback behavior reflection of user is fuzzyyer, a given content recommendation, no matter be positive feedback behavior or negative feedback behavior, be difficult to judge exactly user feedback for specific object, cause the accuracy of user interest of discovery low; On the other hand, often need a large amount of feedback behaviors of user accurately could catch user interest, because this process is usually consuming time longer, be therefore unfavorable for Timeliness coverage user interest.
Summary of the invention
The embodiment of the present invention provides a kind of user interest discover method and device, to improve accuracy and the promptness of the user interest of discovery.
The embodiment of the present invention is by the following technical solutions:
On the one hand, embodiments provide a kind of user interest discover method, comprising:
Obtain the expression data to content recommendation or the behavioral data of user's input;
According to described expression data or described behavioral data, determine that interest predicts the outcome;
Point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
Further, according to described expression data, determine that interest predicts the outcome, comprising:
According to described expression data, determine the object of interest corresponding with described expression data and user's attitude;
According to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.
Further, before the expression data to content recommendation obtaining user's input, also comprise:
Obtain the behavioral data to content recommendation of user's input;
According to described behavioral data, determine the satisfaction of user to described content recommendation;
If described satisfaction is less than setting threshold value, then trigger the operation performing the expression data to content recommendation obtaining user's input.
Further, according to described expression data, determine the object of interest corresponding with described expression data, comprising:
The word that described expression data comprises is mated in object repository;
Using the word that the match is successful as object of interest corresponding to described expression data.
Further, using the word that the match is successful as after the object of interest that described expression data is corresponding, also comprise:
According to the text distance between the word that described object of interest is corresponding with described user's attitude, filter out the object of interest that text distance is greater than setting value.
Further, according to described expression data, determine the user attitude corresponding with described expression data, comprising:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
Further, according to described behavioral data, before determining that interest predicts the outcome, also comprise:
According to described behavioral data, determine the satisfaction of user to described content recommendation;
If described satisfaction is less than setting threshold value, then triggers and perform according to described behavioral data, determine the operation that interest predicts the outcome.
Further, according to described behavioral data, determine the satisfaction of user to described content recommendation, comprise following at least one item:
According to the refreshing frequency of user to content recommendation, determine the satisfaction of user to described content recommendation;
According to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation;
According to user to the support feedback data of content recommendation and the time of concern, determine the satisfaction of user to described content recommendation.
Further, pointing out to applications client described interest to predict the outcome, and obtaining the click data that user predicts the outcome to described interest, after determining user interest, also comprising:
According to the user interest determined, revise the content recommendation pushed to user.
On the other hand, the embodiment of the present invention additionally provides a kind of user interest and finds device, comprising:
User data acquisition module, for obtaining the expression data to content recommendation or the behavioral data of user's input;
Interest predicts the outcome determination module, for according to described expression data or described behavioral data, determines that interest predicts the outcome;
Interest determination module, for pointing out described interest to predict the outcome to applications client, and obtains the click data that user predicts the outcome to described interest, to determine user interest.
Further, the interest determination module that predicts the outcome comprises:
Object of interest determining unit, for according to described expression data, determines the object of interest corresponding with described expression data;
User's attitude determining unit, for according to described expression data, determines the user attitude corresponding with described expression data;
Interest predicts the outcome determining unit, for according to the object of interest corresponding with described expression data and user's attitude, determines that interest predicts the outcome.
Further, described device also comprises:
First satisfaction determination module, for before the expression data to content recommendation obtaining user's input, according to described behavioral data, determines the satisfaction of user to described content recommendation;
Expression data obtains trigger module, during for being less than setting threshold value in described satisfaction, triggers the operation performing the expression data to content recommendation obtaining user's input.
Further, object of interest determining unit comprises:
Coupling subelement, for mating the word that described expression data comprises in object repository;
Object of interest determination subelement, for using the word that the match is successful as object of interest corresponding to described expression data.
Further, object of interest determining unit also comprises:
Object of interest filters subelement, for using the word that the match is successful as after the object of interest that described expression data is corresponding, according to the text distance between the word that described object of interest is corresponding with described user's attitude, filter out the object of interest that text distance is greater than setting value.
Further, user's attitude determining unit specifically for:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
Further, described device also comprises:
Second satisfaction determination module, for according to described behavioral data, before determining that interest predicts the outcome, according to described behavioral data, determines the satisfaction of user to described content recommendation;
Interest predicts the outcome and determines trigger module, if be less than setting threshold value for described satisfaction, then trigger and performs according to described behavioral data, determine that interest predicts the outcome.
Further, the first satisfaction determination module or the second satisfaction determination module comprise following at least one item:
First satisfaction determining unit, for according to the refreshing frequency of user to content recommendation, determines the satisfaction of user to described content recommendation;
Second satisfaction determining unit, for according to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation;
3rd satisfaction determining unit, for according to user to the support feedback data of content recommendation and pay close attention to the time, determine the satisfaction of user to described content recommendation.
Further, described device also comprises:
Pushing module, for pointing out described interest to predict the outcome to applications client, and obtains the click data that user predicts the outcome to described interest, after determining user interest, according to the user interest determined, revises the content recommendation pushed to user.
The Advantageous Effects of the technical scheme that the embodiment of the present invention proposes is: by obtaining the expression data to content recommendation or the behavioral data of user's input, determine that interest predicts the outcome, predict the outcome to applications client reminding interest again, and obtain the click data that user predicts the outcome to interest, to determine user interest.The embodiment of the present invention can allow user initiatively to carry out interest expression by natural language, finds user interest in time, exactly, promotes Consumer's Experience.
Accompanying drawing explanation
In order to be illustrated more clearly in the present invention, introduce doing one to the accompanying drawing used required in the present invention simply below, apparently, accompanying drawing in the following describes is some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the process flow diagram of the user interest discover method that the specific embodiment of the invention one provides;
Fig. 2 is the process flow diagram of the user interest discover method that the specific embodiment of the invention two provides;
Fig. 3 is the process flow diagram of the user interest discover method that the specific embodiment of the invention three provides;
Fig. 4 is the process flow diagram of the user interest discover method that the specific embodiment of the invention four provides;
Fig. 5 is the process flow diagram of the user interest discover method that the specific embodiment of the invention five provides;
Fig. 6 is the process flow diagram of the user interest discover method that the specific embodiment of the invention five provides;
Fig. 7 is the process flow diagram of the user interest discover method that the specific embodiment of the invention five provides;
Fig. 8 is the structured flowchart of the user interest discovery device that the specific embodiment of the invention six provides.
Embodiment
For making the object, technical solutions and advantages of the present invention clearly, be described in further detail the technical scheme in the embodiment of the present invention below in conjunction with accompanying drawing, obviously, described embodiment is the present invention's part embodiment, instead of whole embodiments.Be understandable that; specific embodiment described herein is only for explaining the present invention; but not limitation of the invention; based on the embodiment in the present invention; those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.It also should be noted that, for convenience of description, illustrate only part related to the present invention in accompanying drawing but not full content.
Embodiment one
The process flow diagram of a kind of user interest discover method that Fig. 1 provides for the present embodiment, the method comprises the following steps:
101, the expression data to content recommendation of user's input is obtained.
102, according to described expression data, determine the object of interest corresponding with described expression data, comprising:
The word that described expression data comprises is mated in object repository;
Using the word that the match is successful as object of interest corresponding to described expression data.
103, according to described expression data, determine the user attitude corresponding with described expression data, comprising:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
104, according to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.
105, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
In the present embodiment, expression data specifically can be user and uses natural language to increase, reduce or shield the interest expression data of some content recommendations by hope expressed by voice or text event detection.Such as: " I wants to see more news about 99 formula tanks ", " go out my just unloading of news of hammer mobile phone again! ", " news of millet is too many, you collect money? " etc., user can adopt text input mode or phonetic entry pattern, and text input mode is common input pattern, and no further details to be given herein; And phonetic entry pattern, first the voice of user can be converted to text message by sound identification module, the solution completing a lot of maturation of this step, such as existing speech recognition open platform.As long as recommendation service uses the Freeware development kit of these speech-recognition services business (SDK:Software Development Kit) just freely can utilize the speech-recognition services of these platforms, and this technology is very ripe in practice, recognition accuracy is up to more than 95%.
Identify the intention that user goes out with natural language expressing, its Major Difficulties is to determine corresponding object of interest and user's attitude, i.e. step 102 and step 103 according to expression data.Identification by the intention assessment problem reduction of user is two tuples:
<Object,Attitude>
Wherein Object representative of consumer for object of interest, object of interest is in the above example " 99 formula tank " respectively, " hammer mobile phone ", " millet (millet Science and Technology Ltd.) ".Attitude representative of consumer attitude, can represent with limited set, such as: { ' shielding ', ' reduce and recommend ', ' increase and recommend ' }.
In actual applications, the determination of Object needs dependence Object knowledge base, i.e. an object repository.Object repository needs the significant Object of a large amount of manual confirmation, the title (name) of the another name (ref) corresponding to object of interest, type (type) and object of interest is comprised in knowledge base, wherein type can be classification (category) or entity (entity: name, mechanism's name, makes the concrete 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: scientific and technological >
<ref:[millet, millet science and technology], type:entity, name: millet Science and Technology Ltd. >
<ref:[model ice ice], type:entity, name: model ice ice >
There is object repository, just can map the Object in identification one section words according to another name.
The determination of Attitude mainly according to the rule of some Manual definition, such as:
Give me a little XXX--' less and reduce recommendation '
Recommend some XXX--' to increase to recommend ' more
Not XXX--' shielding '
It should be noted that, under having a kind of situation, certain word in one section of word likely maps multiple Object, such as ' millet ' likely refers to millet Science and Technology Ltd., also the millet eaten likely is referred to, in this case can point out user, to be expressed is any meaning to allow user confirm, or judges which Object is more relevant with user by the use record of user.Such as, if user's expression is negative attitude, the Object appeared in user's browsing histories is more likely the object that user wants to express attitude, confirms that user really wants the Object expressed thus.
Under also having a kind of situation, the Object in the intention of user is not a concrete concept, but compares abstract concept.Such as ' give me a little significant ', the expression that ' not going out vulgar again ' is such.Here ' interesting ' and ' vulgar ' all belongs to and compares abstract concept, cannot correspond to some concrete Object.In this case, can pre-define the label that some are special, such as ' cross-talk ', ' vulgar content ' is as Object.These labels can be marked content by operation personnel, also automatically can identify with machine learning algorithm.Like this, demand more abstract in this expression also can be identified, and then determine Object.
Next, according to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.Point out described interest to predict the outcome to applications client, user selects to click one of them interest and predicts the outcome, and obtains the click data that user predicts the outcome to described interest, to determine user interest.
Such as: above illustrate and " go out my just unloading of news of hammer mobile phone again! ", can determine that Object is " hammer mobile phone ", Attitude is " shielding ", then can predict the outcome as " shielding hammer mobile phone news " supplies user to click to applications client reminding interest, if this predicts the outcome do not meet user's true intention, user can ignore; Or prompting " shielding hammer mobile phone news? ", occur that "Yes" and "No" are to select for user, user selects to click one of them interest and predicts the outcome, and obtains the click data that user predicts the outcome to described interest, to determine user interest.
For another example: above illustrate " I wants to see more news about 99 formula tanks ", can determine that Object is " 99 formula tank ", Attitude is " increase and recommend ", then can predict the outcome as multiple options for user such as " recommending 99 formula tank Recent Progresses In The Development news ", " recommend that 99 formula tanks come into operation situation news " and " recommending western medium about 99 formula tank news " is selected to applications client reminding interest, user selects to click one of them interest and predicts the outcome, obtain the click data that user predicts the outcome to described interest, to determine user interest.
Present embodiments provide a kind of user interest discover method, the method, by obtaining the expression data to content recommendation of user's input, is determined that interest predicts the outcome, then is predicted the outcome to applications client reminding interest, and obtain the click data that user predicts the outcome to interest, to determine user interest.User initiatively can express and understand the interest of oneself to the requirement of content recommendation instead of passive waiting system by very long data mining process, and user can by the demand of natural language expressing oneself, and running cost is very low.The present embodiment can find user interest in time, exactly, promotes Consumer's Experience.
Embodiment two
The process flow diagram of a kind of user interest discover method that Fig. 2 provides for the present embodiment, the method comprises the following steps:
201, the expression data to content recommendation of user's input is obtained.
202, according to described expression data, determine the object of interest corresponding with described expression data, comprising:
The word that described expression data comprises is mated in object repository;
Using the word that the match is successful as object of interest corresponding to described expression data.
203, according to described expression data, determine the user attitude corresponding with described expression data, comprising:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
204, according to the text distance between the described object of interest word corresponding with described user's attitude, the object of interest that text distance is greater than setting value is filtered out.
205, according to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.
206, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
In the present embodiment, expression data specifically can be user and uses natural language to increase, reduce or shield the interest expression data of some content recommendations by hope expressed by voice or text event detection.User can adopt text input mode or phonetic entry pattern.
Identify the intention that user goes out with natural language expressing, its Major Difficulties is to determine corresponding object of interest and user's attitude, i.e. step 202 and step 203 according to expression data.Identification by the intention assessment problem reduction of user is two tuples:
<Object,Attitude>
Wherein Object representative of consumer for object of interest, Attitude representative of consumer attitude, can represent with limited set, such as: { ' shielding ', ' reduce recommend ', ' increase and recommend ' }.
In actual applications, the determination of Object needs dependence Object knowledge base, i.e. an object repository.Object repository needs the significant Object of a large amount of manual confirmation, the title (name) of the another name (ref) corresponding to object of interest, type (type) and object of interest is comprised in knowledge base, wherein type can be classification (category) or entity (entity: name, mechanism's name, makes the concrete things such as the name of an article).There is object repository, just can map the Object in identification one section words according to another name.
The determination of Attitude mainly according to the rule of some Manual definition, such as:
Give me a little XXX--' less and reduce recommendation '
Recommend some XXX--' to increase to recommend ' more
Not XXX--' shielding '
Under having it should be noted that a kind of situation, one section words in can identify multiple Object, such as " millet news is too many, is advertisement? ", " millet news " and " advertisement " in the words can find corresponding Object.In this case, can judge in conjunction with the identification of Attitude, i.e. step 204, if the text distance of an Object and Attitude is greater than setting value, this Object can be filtered out, confirm that user really wants the Object expressed thus, setting value can be that user interest finds that device pre-sets, also can be arranged voluntarily by user, if setting value is " 2 ", when then the text distance of an Object and Attitude is greater than 2, this Object can be filtered out, in upper example " too much " can Attitude be defined as, text distance between " millet news " and " too much " is 0, and the text distance between " advertisement " and " too much " is 3, then can be filled into " advertisement " this Object, confirm that user really wants " the millet news " expressed.
Next, according to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.Point out described interest to predict the outcome to applications client, user selects to click one of them interest and predicts the outcome, and obtains the click data that user predicts the outcome to described interest, to determine user interest.
Present embodiments provide a kind of user interest discover method, the method is on the basis of embodiment one, text distance between adding according to the object of interest word corresponding with user's attitude, filters out the object of interest that text distance is greater than setting value, improves the accuracy rate that object of interest is determined.The present embodiment can find user interest in time, exactly, promotes Consumer's Experience.
Embodiment three
The process flow diagram of a kind of user interest discover method that Fig. 3 provides for the present embodiment, the method comprises the following steps:
301, the behavioral data to content recommendation of user's input is obtained;
302, according to described behavioral data, the satisfaction of user to described content recommendation is determined.Wherein, determine that user can have multiple implementation to the satisfaction of described content recommendation, comprise following at least one item:
According to the refreshing frequency of user to content recommendation, determine the satisfaction of user to described content recommendation;
According to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation;
According to user to the support feedback data of content recommendation and the time of concern, determine the satisfaction of user to described content recommendation.
303, judge whether described satisfaction is less than setting threshold value, if described satisfaction is less than setting threshold value, then triggers the operation performing the expression data to content recommendation obtaining user's input, namely perform step 304.
304, the expression data to content recommendation of user's input is obtained.
305, according to described expression data, determine the object of interest corresponding with described expression data, comprising:
The word that described expression data comprises is mated in object repository;
Using the word that the match is successful as object of interest corresponding to described expression data.
306, according to described expression data, determine the user attitude corresponding with described expression data, comprising:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
307, according to the text distance between the described object of interest word corresponding with described user's attitude, the object of interest that text distance is greater than setting value is filtered out.
308, according to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.
309, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
In the present embodiment, behavioral data specifically can be the refreshing frequency of user to content recommendation, the click data to content recommendation and stay time, to the data corresponding to the behavior in use of the support feedback data and concern time etc. of content recommendation, the satisfaction of user to described content recommendation is determined according to these behavioral datas, setting satisfaction threshold value, if determined satisfaction is higher than threshold value, can continue to recommend, if satisfaction is lower than threshold value, then trigger the operation performing the expression data to content recommendation obtaining user's input.Follow-up operation is identical with embodiment two, and no further details to be given herein.
As further illustrating, in the present embodiment according to the refreshing frequency of user to content recommendation, determine that user to the mode of the satisfaction of described content recommendation can be: when user is unsatisfied with content recommendation, user reads time of content recommendation will be short, and the height of the length of reading time and refreshing frequency is inversely proportional to substantially, the satisfaction corresponding when refreshing frequency height is low, and refreshing frequency low time corresponding satisfaction high, suppose that refreshing frequency be the satisfaction of 5 correspondences per hour is threshold value, when the refreshing frequency of user be 6 times hourly time, the operation performing the expression data to content recommendation obtaining user's input will be triggered,
According to user to the click data of content recommendation and stay time, determine that user to the mode of the satisfaction of described content recommendation can be: adopt grading scheme, article is recommended when user clicks one section, illustrate that user is interested in this article, can score for this content, as 1 point, when user reads this article, reading time is scored, 1 point is remembered as often read 30 seconds, read and be 4 points in 2 minutes, certainly, can adjust obatained score according to article content length, do not describe in detail herein, suppose that accumulative 10 sections of marks recommending article are the score value of satisfaction, threshold value is 20 points, when user have ignored many sections of recommendation articles, satisfaction score value will be very low, when satisfaction score value is lower than 20 timesharing, the operation performing the expression data to content recommendation obtaining user's input will be triggered,
According to user to the support feedback data of content recommendation and the time of concern, determine that user to the mode of the satisfaction of described content recommendation can be: can recommend be provided with customer satisfaction survey option in article and set corresponding score value at every section, if " very satisfied " is 5 points, " feel quite pleased " is 4 points, " generally " is 3 points, " be unsatisfied with " is 2 points, " very dissatisfied " is 1 point, one of them option of user by selecting is this section recommendation essay grade, draw the support feedback data of user to content recommendation, the described concern time can be the concern time of user to the content recommendation in a certain field, as amusement, physical culture and economy etc., the concern time long satisfaction score that then can increase this field content recommendation, last comprehensive above two scores draw final satisfaction score value, when satisfaction score value is lower than threshold value, the operation performing the expression data to content recommendation obtaining user's input will be triggered.These are only the citing of embodiment, be not limited to above embodiment in actual applications.
Present embodiments provide a kind of user interest discover method, the method is on the basis of embodiment two, add the behavioral data to content recommendation obtaining user's input, and according to described behavioral data, determine the step of user to the satisfaction of described content recommendation, if satisfaction is higher, then can reduces the step that user initiatively carries out data representation, promote Consumer's Experience further.
Embodiment four
The process flow diagram of a kind of user interest discover method that Fig. 4 provides for the present embodiment, the method comprises the following steps:
401, the behavioral data to content recommendation of user's input is obtained;
402, according to described behavioral data, the satisfaction of user to described content recommendation is determined.
403, judge whether described satisfaction is less than setting threshold value, if described satisfaction is less than setting threshold value, then trigger the operation performing the expression data to content recommendation obtaining user's input.
404, the expression data to content recommendation of user's input is obtained.
405, according to described expression data, determine the object of interest corresponding with described expression data, comprising:
The word that described expression data comprises is mated in object repository;
Using the word that the match is successful as object of interest corresponding to described expression data.
406, according to described expression data, determine the user attitude corresponding with described expression data, comprising:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
407, according to the text distance between the described object of interest word corresponding with described user's attitude, the object of interest that text distance is greater than setting value is filtered out.
408, according to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.
409, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
410, according to the user interest determined, the content recommendation pushed to user is revised.
Can formulate modification rule for described correction, modification rule can have multiple, and only citing is described herein.Can be such as each section of article setting initial score, adjust accordingly according to the mark of user interest to article, the article that mark is high be more easily recommended, and the low article of mark is more difficult recommended, and the article that mark is zero can conductively-closed.Such as: for user interest < ' millet ', ' reduce and recommend ' >, can trigger " the article mark comprising millet in title reduces by 50% ", " the article mark comprising millet in keyword reduces by 20% " two modification rules.After modification rule comes into force, the article mark of hit " millet " can be punished, thus can be more difficult out recommended.The number of times that the entry-into-force time of modification rule can express according to user is determined.If user expresses first time, the entry-into-force time can be three days, and second time can be one week, and third time can be one month; And for < ' Huawei ', ' increase and recommend ' >, the modification rule of " the article mark comprising Huawei in title doubles " can be triggered, certainly, modification rule also can be other forms, as: " if the article lazy weight 5 sections comprising Huawei in recommended candidate article in title adds recommended candidate from the article database search 5 sections ".In addition, the frequency initiatively can carrying out expressing according to user adjusts modification rule, makes user see the improvement of recommendation results quickly.
Present embodiments provide a kind of user interest discover method, the method is on the basis of embodiment three, add the user interest according to determining, revise the step of the content recommendation pushed to user, this step can make the expression of user come into force, when recommending after expression, user just can see the improvement of recommendation results next time, can promote Consumer's Experience further.
Embodiment five
The process flow diagram of a kind of user interest discover method that Fig. 5 provides for the present embodiment, the method comprises the following steps:
501, the behavioral data to content recommendation of user's input is obtained.
502, according to described behavioral data, determine that interest predicts the outcome.
503, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
In the present embodiment, behavioral data specifically can be the refreshing frequency of user to content recommendation, to click data and the stay time of content recommendation, to the data corresponding to the behavior in use of the support feedback data and concern time etc. of content recommendation, as further illustrating, when user is unsatisfied with content recommendation, user reads time of content recommendation will be short, and the height of the length of reading time and refreshing frequency is inversely proportional to substantially, illustrate that when refreshing frequency height user loses interest in content recommendation, and refreshing frequency low time illustrate user interested in content recommendation, determine that interest predicts the outcome for user's selection with this, when user click one section recommend article, illustrates that user is interested in this article, when user read this article spended time long time, very interested in this article, with this determine interest predict the outcome for user selection, can recommend to be provided with customer satisfaction survey option in article at every section in addition, as " very satisfied ", " feeling quite pleased ", " ", " being unsatisfied with " and " very dissatisfied ", one of them option of user by selecting, draw the support feedback data of user to content recommendation, the described concern time can be the concern time of user to the content recommendation in a certain field, as amusement, physical culture and economy etc., the concern time, long then explanation user was interested in this field content recommendation, finally determined that interest predicts the outcome for user's selection.These are only the citing of embodiment, be not limited to above embodiment in actual applications.
Such as: if user have ignored the many sections of recommendation articles about millet, so user probably loses interest in millet.If user is more to the article number of clicks about electric automobile, so user is probably also interested in the news of Tesla (CS) Koncern, Podebradska 186, Praha 9, Czechoslovakia founder Maas gram.Traditionally, these information can be used directly to reduce or more recommendation class news.But because the behavior of user is complicated, this prediction is likely wrong.The satisfaction that probably can not promote user according to these prediction recommendation articles blindly.So, according to described behavioral data, can determine that interest predicts the outcome, predict the outcome to applications client reminding interest, and obtain the click data that user predicts the outcome to described interest, to determine user interest.Such as, can predict the outcome to the following interest of Client-Prompt,
Whether you think:
Reduce the recommendation about millet
Recommend the news of Maas gram
Shield the news of net of cutting firewood
If have a wish meeting user in predicting the outcome above-mentioned, user can click corresponding interest and predict the outcome to determine user interest.
Present embodiments provide a kind of user interest discover method, the method, according to user behavior data, determines that interest predicts the outcome, and points out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.The present embodiment can predict according to user behavior data the recommended requirements that user is possible, once predict successfully, user only needs to confirm, can save voice or text event detection step, can promote Consumer's Experience further.
Embodiment six
The process flow diagram of a kind of user interest discover method that Fig. 6 provides for the present embodiment, the method comprises the following steps:
601, the behavioral data to content recommendation of user's input is obtained.
602, according to described behavioral data, the satisfaction of user to described content recommendation is determined.
603, judge whether described satisfaction is less than setting threshold value, if so, then trigger performing according to described behavioral data, determine the operation that interest predicts the outcome, namely perform step 604.
604, according to described behavioral data, determine that interest predicts the outcome.
605, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
In the present embodiment, behavioral data specifically can be the refreshing frequency of user to content recommendation, the click data to content recommendation and stay time, to the data corresponding to the behavior in use of the support feedback data and concern time etc. of content recommendation, the satisfaction of user to described content recommendation is determined according to these behavioral datas, setting satisfaction threshold value, if determined satisfaction is higher than threshold value, can continue to recommend, if satisfaction is lower than threshold value, then trigger and perform according to described behavioral data, determine that interest predicts the outcome.
Present embodiments provide a kind of user interest discover method, the method is according to behavioral data, determine the satisfaction of user to content recommendation, if satisfaction is less than setting threshold value, then trigger and perform according to behavioral data, determine that interest predicts the outcome, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.The number of times that user initiatively carries out the step of data representation can be reduced, greatly reduce use cost, Consumer's Experience can be promoted further simultaneously.
Embodiment seven
The process flow diagram of a kind of user interest discover method that Fig. 7 provides for the present embodiment, the method comprises the following steps:
701, the behavioral data to content recommendation of user's input is obtained.
702, according to described behavioral data, determine that interest predicts the outcome.
703, point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
704, according to the user interest determined, the content recommendation pushed to user is revised.
The present embodiment adds the user interest according to determining on the basis of embodiment five, revises this step of content recommendation pushed to user, can promote Consumer's Experience further.
Embodiment eight
Based on the various embodiments described above, the present embodiment provides a kind of user interest to find device, and as shown in Figure 8, be the structured flowchart of the present embodiment device, this device can comprise:
User data acquisition module 801, for obtaining the expression data to content recommendation or the behavioral data of user's input.
Interest predicts the outcome determination module 802, for according to described expression data or described behavioral data, determines that interest predicts the outcome, comprising: object of interest determining unit, for according to described expression data, determine the object of interest corresponding with described expression data; User's attitude determining unit, for according to described expression data, determines the user attitude corresponding with described expression data; Interest predicts the outcome determining unit, for according to the object of interest corresponding with described expression data and user's attitude, determines that interest predicts the outcome.Wherein, object of interest determining unit specifically comprises: the first coupling subelement, for mating the word that described expression data comprises in object repository; Object of interest determination subelement, for using the word that the match is successful as object of interest corresponding to described expression data; Object of interest filters subelement, for using the word that the match is successful as after the object of interest that described expression data is corresponding, according to the text distance between the word that described object of interest is corresponding with described user's attitude, filter out the object of interest that text distance is greater than setting value.Wherein, user's attitude determining unit comprises: the second coupling subelement, for mating the word that described expression data comprises in default emotional attitude template; User's attitude determination subelement, for according to matching result, determines the user attitude corresponding with described expression data.
Interest determination module 803, for pointing out described interest to predict the outcome to applications client, and obtains the click data that user predicts the outcome to described interest, to determine user interest.
First satisfaction determination module 804, for before the expression data to content recommendation obtaining user's input, according to described behavioral data, determine the satisfaction of user to described content recommendation, comprise following at least one item: the first satisfaction determining unit, for according to the refreshing frequency of user to content recommendation, determine the satisfaction of user to described content recommendation; Second satisfaction determining unit, for according to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation; 3rd satisfaction determining unit, for according to user to the support feedback data of content recommendation and pay close attention to the time, determine the satisfaction of user to described content recommendation.
Expression data obtains trigger module 805, if be less than setting threshold value for described satisfaction, then triggers the operation performing the expression data to content recommendation obtaining user's input.
Second satisfaction determination module 806, for according to described behavioral data, before determining that interest predicts the outcome, according to described behavioral data, determine the satisfaction of user to described content recommendation, comprise following at least one item: the first satisfaction determining unit, for according to the refreshing frequency of user to content recommendation, determine the satisfaction of user to described content recommendation; Second satisfaction determining unit, for according to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation; 3rd satisfaction determining unit, for according to user to the support feedback data of content recommendation and pay close attention to the time, determine the satisfaction of user to described content recommendation.
Interest predicts the outcome and determines trigger module 807, if be less than setting threshold value for described satisfaction, then trigger and performs according to described behavioral data, determine that interest predicts the outcome.
Pushing module 808, for pointing out described interest to predict the outcome to applications client, and obtains the click data that user predicts the outcome to described interest, after determining user interest, according to the user interest determined, revises the content recommendation pushed to user.
A kind of user interest discovery device that the embodiment of the present invention provides can perform the user interest discover method that any embodiment of the present invention provides, and possesses the corresponding functional module of manner of execution and beneficial effect.
Last it is noted that above each embodiment is only for illustration of technical scheme of the present invention, but not be limited; In embodiment preferred embodiment, be not limited, to those skilled in the art, the present invention can have various change and change.All do within spirit of the present invention and principle any amendment, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (18)
1. a user interest discover method, is characterized in that, comprising:
Obtain the expression data to content recommendation or the behavioral data of user's input;
According to described expression data or described behavioral data, determine that interest predicts the outcome;
Point out described interest to predict the outcome to applications client, and obtain the click data that user predicts the outcome to described interest, to determine user interest.
2. method according to claim 1, is characterized in that, according to described expression data, determines that interest predicts the outcome, comprising:
According to described expression data, determine the object of interest corresponding with described expression data and user's attitude;
According to the object of interest corresponding with described expression data and user's attitude, determine that interest predicts the outcome.
3. method according to claim 2, is characterized in that, before the expression data to content recommendation obtaining user's input, also comprises:
Obtain the behavioral data to content recommendation of user's input;
According to described behavioral data, determine the satisfaction of user to described content recommendation;
If described satisfaction is less than setting threshold value, then trigger the operation performing the expression data to content recommendation obtaining user's input.
4. method according to claim 2, is characterized in that, according to described expression data, determines the object of interest corresponding with described expression data, comprising:
The word that described expression data comprises is mated in object repository;
Using the word that the match is successful as object of interest corresponding to described expression data.
5. method according to claim 4, is characterized in that, using the word that the match is successful as after the object of interest that described expression data is corresponding, also comprise:
According to the text distance between the word that described object of interest is corresponding with described user's attitude, filter out the object of interest that text distance is greater than setting value.
6. method according to claim 2, is characterized in that, according to described expression data, determines the user attitude corresponding with described expression data, comprising:
The word that described expression data comprises is mated in default emotional attitude template;
According to matching result, determine the user attitude corresponding with described expression data.
7. method according to claim 1, is characterized in that, according to described behavioral data, before determining that interest predicts the outcome, also comprises:
According to described behavioral data, determine the satisfaction of user to described content recommendation;
If described satisfaction is less than setting threshold value, then triggers and perform according to described behavioral data, determine the operation that interest predicts the outcome.
8. the method according to claim 3 or 7, is characterized in that, according to described behavioral data, determines the satisfaction of user to described content recommendation, comprises following at least one item:
According to the refreshing frequency of user to content recommendation, determine the satisfaction of user to described content recommendation;
According to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation;
According to user to the support feedback data of content recommendation and the time of concern, determine the satisfaction of user to described content recommendation.
9. according to the arbitrary described method of claim 1-7, it is characterized in that, pointing out to applications client described interest to predict the outcome, and obtain the click data that user predicts the outcome to described interest, after determining user interest, also comprise:
According to the user interest determined, revise the content recommendation pushed to user.
10. user interest finds a device, it is characterized in that, comprising:
User data acquisition module, for obtaining the expression data to content recommendation or the behavioral data of user's input;
Interest predicts the outcome determination module, for according to described expression data or described behavioral data, determines that interest predicts the outcome;
Interest determination module, for pointing out described interest to predict the outcome to applications client, and obtains the click data that user predicts the outcome to described interest, to determine user interest.
11. devices according to claim 10, is characterized in that, the interest determination module that predicts the outcome comprises:
Object of interest determining unit, for according to described expression data, determines the object of interest corresponding with described expression data;
User's attitude determining unit, for according to described expression data, determines the user attitude corresponding with described expression data;
Interest predicts the outcome determining unit, for according to the object of interest corresponding with described expression data and user's attitude, determines that interest predicts the outcome.
12. devices according to claim 11, is characterized in that, described device also comprises:
First satisfaction determination module, for before the expression data to content recommendation obtaining user's input, according to described behavioral data, determines the satisfaction of user to described content recommendation;
Expression data obtains trigger module, during for being less than setting threshold value in described satisfaction, triggers the operation performing the expression data to content recommendation obtaining user's input.
13. devices according to claim 11, is characterized in that, object of interest determining unit comprises:
First coupling subelement, for mating the word that described expression data comprises in object repository;
Object of interest determination subelement, for using the word that the match is successful as object of interest corresponding to described expression data.
14. devices according to claim 13, is characterized in that, object of interest determining unit also comprises:
Object of interest filters subelement, for using the word that the match is successful as after the object of interest that described expression data is corresponding, according to the text distance between the word that described object of interest is corresponding with described user's attitude, filter out the object of interest that text distance is greater than setting value.
15. devices according to claim 11, is characterized in that, user's attitude determining unit comprises:
Second coupling subelement, for mating the word that described expression data comprises in default emotional attitude template;
User's attitude determination subelement, for according to matching result, determines the user attitude corresponding with described expression data.
16. devices according to claim 10, is characterized in that, described device also comprises:
Second satisfaction determination module, for according to described behavioral data, before determining that interest predicts the outcome, according to described behavioral data, determines the satisfaction of user to described content recommendation;
Interest predicts the outcome and determines trigger module, during for being less than setting threshold value in described satisfaction, triggering and performs according to described behavioral data, determine the operation that interest predicts the outcome.
17. devices according to claim 12, is characterized in that, the first satisfaction determination module comprises following at least one item:
First satisfaction determining unit, for according to the refreshing frequency of user to content recommendation, determines the satisfaction of user to described content recommendation;
Second satisfaction determining unit, for according to user to the click data of content recommendation and stay time, determine the satisfaction of user to described content recommendation;
3rd satisfaction determining unit, for according to user to the support feedback data of content recommendation and pay close attention to the time, determine the satisfaction of user to described content recommendation.
18. according to the arbitrary described device of claim 10-17, and it is characterized in that, described device also comprises:
Pushing module, for pointing out described interest to predict the outcome to applications client, and obtains the click data that user predicts the outcome to described interest, after determining user interest, according to the user interest determined, revises the content recommendation pushed to user.
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 true CN104361063A (en) | 2015-02-18 |
CN104361063B 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) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105303398A (en) * | 2015-09-29 | 2016-02-03 | 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 |
CN106850762A (en) * | 2017-01-03 | 2017-06-13 | 腾讯科技(深圳)有限公司 | A kind of information push method, server and message push system |
CN107273489A (en) * | 2017-06-14 | 2017-10-20 | 掌阅科技股份有限公司 | Content delivery method, electronic equipment and computer-readable storage medium |
CN107766467A (en) * | 2017-09-29 | 2018-03-06 | 北京金山安全软件有限公司 | Information detection method and device, electronic equipment and storage medium |
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 |
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 |
CN107911491A (en) * | 2017-12-27 | 2018-04-13 | 广东欧珀移动通信有限公司 | Information recommendation method, device and storage medium, server and mobile terminal |
CN108305091A (en) * | 2017-01-13 | 2018-07-20 | 北京京东尚科信息技术有限公司 | Electronic equipment, user's perceptibility extracting method interested and device |
CN108431811A (en) * | 2015-12-23 | 2018-08-21 | 三星电子株式会社 | The method and its electronic device of content are provided a user according to the preference of user |
CN108733666A (en) * | 2017-04-13 | 2018-11-02 | 腾讯科技(深圳)有限公司 | Server info method for pushing, end message sending method and device, system |
CN108985548A (en) * | 2017-06-05 | 2018-12-11 | 埃森哲环球解决方案有限公司 | Real-time intelligent and dynamic delivering arrange |
CN110516159A (en) * | 2019-08-30 | 2019-11-29 | 北京字节跳动网络技术有限公司 | A kind of information recommendation method, device, electronic equipment and storage medium |
CN111444438A (en) * | 2020-03-24 | 2020-07-24 | 北京百度网讯科技有限公司 | Method, device, equipment and storage medium for determining recall permission rate of recall strategy |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070239535A1 (en) * | 2006-03-29 | 2007-10-11 | Koran Joshua M | Behavioral targeting system that generates user profiles for target objectives |
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 |
-
2014
- 2014-11-04 CN CN201410613040.3A patent/CN104361063B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070239535A1 (en) * | 2006-03-29 | 2007-10-11 | Koran Joshua M | Behavioral targeting system that generates user profiles for target objectives |
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 |
Cited By (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105303398A (en) * | 2015-09-29 | 2016-02-03 | 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 |
CN108431811A (en) * | 2015-12-23 | 2018-08-21 | 三星电子株式会社 | The method and its electronic device of content are provided a user according to the preference of user |
US10956473B2 (en) | 2016-09-22 | 2021-03-23 | Guangzhou Ucweb Computer Technology Co., Ltd. | Article quality scoring method and device, client, server, and programmable device |
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 |
CN106850762A (en) * | 2017-01-03 | 2017-06-13 | 腾讯科技(深圳)有限公司 | A kind of information push method, server and message push system |
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 |
CN108305091A (en) * | 2017-01-13 | 2018-07-20 | 北京京东尚科信息技术有限公司 | Electronic equipment, user's perceptibility extracting method interested 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 |
CN108733666A (en) * | 2017-04-13 | 2018-11-02 | 腾讯科技(深圳)有限公司 | Server info method for pushing, end message sending method and device, system |
US11270246B2 (en) | 2017-06-05 | 2022-03-08 | Accenture Global Solutions Limited | Real-time intelligent and dynamic delivery scheduling |
CN108985548A (en) * | 2017-06-05 | 2018-12-11 | 埃森哲环球解决方案有限公司 | Real-time intelligent and dynamic delivering arrange |
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 |
CN107273489A (en) * | 2017-06-14 | 2017-10-20 | 掌阅科技股份有限公司 | Content delivery method, electronic equipment and computer-readable 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 |
CN107766467A (en) * | 2017-09-29 | 2018-03-06 | 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 |
CN107911491A (en) * | 2017-12-27 | 2018-04-13 | 广东欧珀移动通信有限公司 | Information recommendation method, device and storage medium, server and mobile terminal |
CN110516159A (en) * | 2019-08-30 | 2019-11-29 | 北京字节跳动网络技术有限公司 | A kind of information recommendation method, device, electronic equipment and storage medium |
CN111444438A (en) * | 2020-03-24 | 2020-07-24 | 北京百度网讯科技有限公司 | Method, device, equipment and storage medium for determining recall permission rate of recall strategy |
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 |
Also Published As
Publication number | Publication date |
---|---|
CN104361063B (en) | 2018-03-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104361063A (en) | User interest discovering method and device | |
CN108647205B (en) | Fine-grained emotion analysis model construction method and device and readable storage medium | |
CN1934569B (en) | Search systems and methods with integration of user annotations | |
US9256761B1 (en) | Data storage service for personalization system | |
CN103098051B (en) | Search engine optmization assistant | |
US20120290509A1 (en) | Training Statistical Dialog Managers in Spoken Dialog Systems With Web Data | |
US10902341B1 (en) | Machine learning based list recommendations | |
CN111144723A (en) | Method and system for recommending people's job matching and storage medium | |
CN106462565A (en) | Updating text within a document | |
CN111625632A (en) | Question-answer pair recommendation method, device, equipment and storage medium | |
CN106033466A (en) | Database query method and device | |
WO2018081020A1 (en) | Computerized domain expert | |
CN110674312B (en) | Method, device and medium for constructing knowledge graph and electronic equipment | |
CN103995870A (en) | Interactive searching method and device | |
CN101004737A (en) | Individualized document processing system based on keywords | |
CN104572962A (en) | APP (Application) recommendation method and system | |
CN106688215A (en) | Automated click type selection for content performance optimization | |
CN107861753B (en) | APP generation index, retrieval method and system and readable storage medium | |
EP2707807A2 (en) | Training statistical dialog managers in spoken dialog systems with web data | |
CN105117380A (en) | Paste processing method and device | |
CN102314452A (en) | Method for navigation through input method platform and system | |
CN105095311A (en) | Method, device and system for processing promotion information | |
CN114860916A (en) | Knowledge retrieval method and device | |
CN114817755A (en) | User interaction content management method, device and storage medium | |
CN111931077A (en) | Data processing method and device, electronic equipment and storage medium |
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 |