CN107831915A - One kind input complementing method, device, electronic equipment and readable storage medium storing program for executing - Google Patents
One kind input complementing method, device, electronic equipment and readable storage medium storing program for executing Download PDFInfo
- Publication number
- CN107831915A CN107831915A CN201710964623.4A CN201710964623A CN107831915A CN 107831915 A CN107831915 A CN 107831915A CN 201710964623 A CN201710964623 A CN 201710964623A CN 107831915 A CN107831915 A CN 107831915A
- Authority
- CN
- China
- Prior art keywords
- input
- content
- completion
- user
- input content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/02—Input arrangements using manually operated switches, e.g. using keyboards or dials
- G06F3/023—Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
- G06F3/0233—Character input methods
- G06F3/0236—Character input methods using selection techniques to select from displayed items
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- User Interface Of Digital Computer (AREA)
- Stored Programmes (AREA)
Abstract
The invention provides one kind input complementing method, device, electronic equipment and readable storage medium storing program for executing, methods described to include：If detecting input operation of the user in the page is set, input content contextual information corresponding to the input operation is obtained；The contextual information includes the contextual information of the setting page；According to the input content and the contextual information, using default completion model, the completion content for meeting user view is obtained；The completion content, which carries, recommends user.Solves the candidate word for inputting association in first technology according only to user, user view can not be embodied, the problem of reducing the enthusiasm and liveness and comment quality of user, intelligent completion user input can be carried out by contextual information, improves the enthusiasm and liveness and comment quality of user.
Description
Technical field
The present embodiments relate to field of computer technology, more particularly to a kind of input complementing method, device, electronic equipment
And readable storage medium storing program for executing.
Background technology
With the development of network, people can pass through Online Shopping extensive stock and service.In above process, network needs
To be retrieved according to the words and phrases that user inputs.In order to improve user's input efficiency, during user inputs words and phrases, client
The words and phrases association inputted according to user goes out some options for user selections.For example, user can to commodity and service into
Row comment when, when user input " speed " when, client can association go out " fast " or it is " slow " supply user selection.
At present, input association's scheme goes out candidate word all in accordance with history input record and the current input word association of user.It is this
Scheme records each input word of user, according to current input word, obtains candidate word by keyword retrieval and recommends
User.
But the candidate word of association is inputted according only to user, it is impossible to which the comment for embodying user completely is intended to, and reduces user's
Enthusiasm and liveness and comment quality.
The content of the invention
The present invention provides a kind of input complementing method, device, electronic equipment and readable storage medium storing program for executing, to solve prior art
The related above mentioned problem of middle input completion.
According to the first aspect of the invention, there is provided one kind input complementing method, methods described include：
If detecting input operation of the user in the page is set, input content corresponding to the input operation is obtained
Contextual information；The contextual information includes the contextual information of the setting page；
According to the input content and the contextual information, using default completion model, acquisition meets user view
Completion content；
By the completion commending contents to user.
According to the second aspect of the invention, there is provided one kind input complementing device, described device include：
Data obtaining module, if for detecting input operation of the user in the page is set, obtain the input
Input content contextual information corresponding to operation；The contextual information includes the contextual information of the setting page；
Completion content obtaining module, for according to the input content and the contextual information, utilizing default completion
Model, obtain the completion content for meeting user view；
Recommending module, for by the completion commending contents to user.
According to the third aspect of the invention we, there is provided a kind of electronic equipment, including：
Processor, memory and it is stored in the computer journey that can be run on the memory and on the processor
Sequence, it is characterised in that foregoing input complementing method is realized during the computing device described program.
According to the fourth aspect of the invention, there is provided a kind of readable storage medium storing program for executing, it is characterised in that when the storage medium
In instruction by electronic equipment computing device when so that electronic equipment is able to carry out foregoing input complementing method.
The embodiments of the invention provide one kind input complementing method, device, electronic equipment and readable storage medium storing program for executing, the side
Method includes：If detecting input operation of the user in the page is set, input content corresponding to the input operation is obtained
Contextual information；The contextual information includes the contextual information of the setting page；According to the input content and described
Contextual information, using default completion model, obtain the completion content for meeting user view；The completion content, which carries, to be recommended
User.Solves the candidate word for inputting association in first technology according only to user, it is impossible to embody user view, reduce the product of user
The problem of polarity and liveness and comment quality, can by the contextual information such as contextual information including setting the page,
Intelligent completion is carried out to input operation of the user in the page is set, improves the enthusiasm and liveness and comment matter of user
Amount.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below by institute in the description to the embodiment of the present invention
The accompanying drawing needed to use is briefly described, it should be apparent that, drawings in the following description are only some implementations of the present invention
Example, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these accompanying drawings
Obtain other accompanying drawings.
Fig. 1 is a kind of input complementing method specific steps flow chart under system architecture provided in an embodiment of the present invention；
Figure 1A is a kind of review pages example of the embodiment of the present invention；
Figure 1B is the inputting interface example without recommendation process of the present invention based on Figure 1A；
Fig. 1 C are the inputting interface examples without recommendation process of the present invention based on Figure 1A；
Fig. 1 D are a kind of review pages examples of the embodiment of the present invention；
Fig. 1 E are the inputting interface examples without recommendation process of the present invention based on Fig. 1 D；
Fig. 1 F are the inputting interface examples without recommendation process of the present invention based on Fig. 1 D；
Fig. 2 is another input complementing method specific steps flow chart under system architecture provided in an embodiment of the present invention；
Fig. 2A is that the completion model of the embodiment of the present invention shows that training process is intended to；
Fig. 3 is a kind of structure chart for inputting complementing device provided in an embodiment of the present invention；
Fig. 4 is the structure chart of another input complementing device provided in an embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is part of the embodiment of the present invention, rather than whole embodiments.Based on this hair
Embodiment in bright, the every other implementation that those of ordinary skill in the art are obtained under the premise of creative work is not made
Example, belongs to the scope of protection of the invention.
Embodiment one
Reference picture 1, it illustrates a kind of step flow chart of mobile input complementing method, including：
Step 101, if detecting input operation of the user in the page is set, obtain corresponding to the input operation
Input content contextual information；The contextual information includes the contextual information of the setting page.
The embodiment of the present invention is applied to the application with input function, such as：Search engine, there is answering for comment function
With, input method etc..
Wherein, it can be review pages to set the page.Such as Figure 1A, it illustrates the setting page of a Peking Duck Restaurant to show
Example, wherein have the marking item of five stars, user's most 5 stars of skyer, minimum optional 0 star；Then user can be defeated in comment
Enter and comment is inputted in frame.Certainly, for the setting page, the also marking of sundry item, such as speed, environment etc..
The input operation detected in the page is set can be by detecting dead time or user in user's input process
Input the modes such as predetermined symbol.For example, detect that the dead time is more than preset time threshold, or detection user inputs default sky
Lattice key, comma or other designated symbols.The embodiment of the present invention is not any limitation as to it.
It should be noted that the contextual information of the setting page can include user in the page is set to projects
Marking, such as foregoing total score, environment, speed etc., can be the comment object that the setting page is directed to, such as in Figure 1A
" * * * " Peking Duck Restaurants, the comment relevant information for recommending input inside the non-comment input frame such as dish can also be included.Certainly, inputting
During the contextual information can also include user in input, the content that has inputted, for example, user have input it is interior
After holding " speed ", currently to continue to input, it is exactly the contextual information inputted to be somebody's turn to do " speed ".That is, the setting page
Contextual information includes the information associated with institute input content.
Step 102, according to the input content and the contextual information, using default completion model, acquisition meets
The completion content of user view.
In embodiments of the present invention, completion model can be trained to obtain according to sample content, and the sample content includes each
User is in the history input content of the setting page and the contextual information of the setting page.
In one alternatively scheme, completion model can include maker, determining device, and maker can generate the sample of simulation
This content, determining device are judged the sample content of simulation.When determining device cannot be distinguished by simulation and real sample content,
Training terminates, and shows that maker can generate intimate real sample content；Otherwise, the parameter of maker and determining device is adjusted,
Continue training until determining device cannot be distinguished by simulation and real sample content.
In this way, according to input content and contextual information, the completion model is inputted, the benefit for meeting user view can be obtained
Full content.For example user has beaten the star of total score one, it have input " environment " two word, can be commented based on the difference of a star, give prediction " noise
It is miscellaneous " two words.
Step 103, by the completion commending contents to user.
Specifically, completion content is shown on screen, for selection by the user.It is for example, completion content is currently displayed
Set interface specified location or be shown in the candidate item of input method.When completion content includes multiple, it is necessary to according to one
Fixed order is shown.It is appreciated that the embodiment of the present invention is not any limitation as to display location and size, form etc..
The lower several recommendation process for illustrating the embodiment of the present invention of mask.
Such as Figure 1B, it is the input method candidate for the recommendation process for not carrying out embodiment of the present invention on the basis of Figure 1A
Item examples of interfaces.
User clicks on Figure 1A comment input frame, and into Figure 1B comment inputting interface, then user have input " environment ".
Do not use the embodiment of the present invention recommendation process then preference it is recommended that " much better ".
Such as Fig. 1 C, it is candidates of input method circle after the recommendation process on the basis of Figure 1A using the embodiment of the present invention
Face example.
User clicks on Figure 1A comment input frame, and into Fig. 1 C comment inputting interface, then user have input " environment ".
Step 110 of the embodiment of the present invention then detects the input operation, then obtains its contextual information, such as " star of total score one ", " environment
The contextual information of the review pages such as difference ", it can also include " environment " that user has inputted, then pushing away by step 120
Process is recommended, can recommend " noisy ".Then " noisy " is put into input method preference by step 130.
Certainly, its completion content recommended can also be illustrated in the specified location of review pages.Such as comment frame side etc.
Position.
Another example is enumerated again：
Fig. 1 D show another review pages example.With the marking that 5 stars are given to total score.It has selected and recommend dish roast duck.
Such as Fig. 1 E, it is the input method candidate for the recommendation process for not carrying out embodiment of the present invention on the basis of Fig. 1 D
Item examples of interfaces.
User clicks on Fig. 1 D comment input frame, and into Fig. 1 E comment inputting interface, then user have input that " taste is very
It is good, especially ".Do not use the embodiment of the present invention recommendation process then preference it is recommended that " you ".
Such as Fig. 1 F, it is candidates of input method circle after the recommendation process on the basis of Figure 1A using the embodiment of the present invention
Face example.
User clicks on Figure 1A comment input frame, and into Fig. 1 F comment inputting interface, then user have input " environment ".
Step 110 of the embodiment of the present invention then detects the input operation, then obtains its contextual information, such as " total score five-pointed star ", " service
Well ", " dish is recommended：The contextual information of the review pages such as roast duck ", can also have been inputted including user " it is delicious, especially
It is ", then by the recommendation process of step 120, can recommend " roast duck ".Then " roast duck " is put into input method by step 130
Preference.
In summary, the embodiments of the invention provide one kind to input complementing method, and methods described includes：If detect use
Input operation of the family in the page is set, then obtain input content contextual information corresponding to the input operation；Above and below described
Literary information includes the contextual information of the setting page；According to the input content and the contextual information, using default
Completion model, obtain and meet the completion content of user view；The completion content, which carries, recommends user.Solves first technology
In the candidate word of association is inputted according only to user, it is impossible to embody user view, reduce the enthusiasm of user and liveness and comment
The problem of by quality, can be by contextual informations such as the contextual informations including setting the page, to user in the page is set
Input operation carry out intelligent completion, improve the enthusiasm and liveness and comment quality of user.
Embodiment two
The embodiment of the present application is described from the level of system architecture to alternatively inputting complementing method.
Reference picture 2, it illustrates the specific steps flow chart of another input complementing method.
Step 201, completion model is trained according to user's sample data；User's sample data includes：Input content, with
The contextual information of the setting page belonging to input content.
The embodiment of the present invention trains completion model by a large number of users sample data, specifically trains the life in completion model
Grow up to be a useful person and determining device, so that completion model can be generated in terms of sentence smoothness degree, the theme degree of accuracy, the scoring degree of accuracy
Qualified sentence.
It is appreciated that user's sample data can include：Input content, the context with the affiliated setting page of input content
Information.By taking user's sample data of comment as an example, it is recorded for the historical review of user, including user comment content, comment master
Topic, scoring to projects etc..Wherein, input content is user comment content, comment theme and score for input content institute
The contextual information of the category setting page.For example, for a five-pointed star comment " dumpling of today is good delicious " on dumpling, comment
Theme is " dumpling ", is scored as " five-pointed star ", and comment content is " dumpling of today is very delicious ".
Certainly, the division of contextual information is also needed to be set according to practical application scene, and the embodiment of the present invention is not added with to it
With limitation.
Alternatively, in another embodiment of the invention, step 201 includes sub-step 2011 to 2014：
Sub-step 2011, the true input content for meeting preparatory condition is selected from user's sample data.
Preparatory condition is user's input content；For example, when user inputs " yellow braised chicken ", master is selected from sample data
The true input content of entitled " yellow braised chicken ".In another example when user have input " star " evaluation, scoring is selected from sample data
For the true input content of " star ".Certainly, preparatory condition can include two kinds of theme and scoring simultaneously.In actual applications,
Preparatory condition can also include other forms, and the embodiment of the present invention is not any limitation as to it.
It is appreciated that different themes and emotion can correspond to corresponding preparatory condition, it is corresponding to select sample to be trained.
In embodiments of the present invention, every group of preparatory condition is all trained.
Sub-step 2012, using the completion model maker according to the preparatory condition generation corresponding to simulation input
Content.
Specifically, maker generates simulation input content according to preparatory condition at random, yellow braised chicken and each word is carried out random
Combination.For example, user's input, " during yellow braised chicken ", maker is generated in a series of simulation inputs on yellow braised chicken according to yellow braised chicken
Hold, including：Yellow braised chicken taste is delicious, and yellow braised chicken is peppery, and yellow braised chicken is fast etc..It will be appreciated, of course, that simulation input content includes one
Logical content is not inconsistent a bit, can be judged by sentence smoothness degree, theme accuracy, scoring accuracy.
Its preparatory condition can be set according to demand.
It is appreciated that needing to initialize maker before using maker, the initial metaplasia of initial weight is used
Grow up to be a useful person.
Sub-step 2013, for the true input content and each input content of simulation input content, using described
The determining device of completion model is judged for the first time, obtains corresponding intermediateness input content.
Specifically, as shown in Figure 2 A, the determining device on the dotted line left side is to each of true input content and simulation input content
Input content carries out probability calculation, obtains the probability of intermediateness input content and the input content for true input content,
When probability reaches 50 percent, it is believed that arbiter cannot be distinguished by true and simulation input content, terminate so as to train；It is no
Then, intensive training is continued.It is appreciated that intermediateness input content is the first stage to be directed to true input content institute to maker
After one simulation input content of generation, the simulation after being differentiated by the arbiter of first stage to the simulation input content is defeated
Enter content, it is possible to understand that intermediateness input content or a sentence, the sentence are the inputs as second stage training.
Sub-step 2014, the maker and the arbiter are strengthened again using the intermediateness input content
Training, obtains the completion model.
In embodiments of the present invention, training completion model includes two stages, and the first stage is initial treatment, corresponding sub-step
Rapid 2013, such as dotted line left-hand component in Fig. 2A, second stage is intensive training, corresponding sub-step 2014, as dotted line is right in Fig. 2A
Rim portion.Wherein, when second stage gives intensive training, it is necessary to constantly change model parameter.Wherein, Fig. 2A first stage
In：G represents maker, and D represents arbiter, and Real word represent real user's sample data；True data represent user
True input content in sample data；Generate represents the simulation input content for True data generations；In the rank
Section G only generates once completely simulation sentence for a true input content.In second stage：G represents maker, and D is represented
Arbiter, it is meant that the first stage is with the maker of second stage as arbiter parameter；Exported for the D of first stage
Sentence, it is considered a state (state) of second stage, and then the G of second stage is performed multiple, generates multiple complete moulds
Intend sentence；Then to differentiating in each simulation input by sentence D, and then Utilization strategies boosting algorithm (policy
Gradient G and D model parameter) is corrected.Circulation proceeds as described above to True data and Generate, is generated until to G
The sentence for meeting aforementioned condition, D identification it is out not true and false, completion model is that foundation finishes.
Alternatively, in another embodiment of the invention, sub-step 2014 includes sub-step 20141 to 20144：
Sub-step 20141, the contextual information based on the intermediateness input content and corresponding true input content,
Multiple first candidate input contents are generated using the maker.
Specifically, it is upper based on intermediateness input content and corresponding true input content according to certain predetermined condition
Context information random combine, generate multiple first candidate input contents.
For example, when intermediateness input content is " dumpling of today ", correspondingly the contextual information of true input content is
Evaluate " star ", theme is " dumpling ", then first candidate's input content can be " dumpling of today is eaten very well ", " dumpling of today
Son be difficult eat ", " dumpling is a lot ", " dumpling is seldom ".
Sub-step 20142, each first candidate input content sample data is judged using the determining device, obtained
Each second candidate input content sample data and corresponding actual value.
Wherein, determining device is a riffle, can be with by constantly carrying out two classification to each candidate's input content sample data
Obtain the probability that it is true input content, i.e. actual value.
When actual value is closer to 50 percent, show that candidate's input content sample data is accurate in sentence smoothness degree, theme
Meet full-scale condition in exactness, the scoring degree of accuracy.
Sub-step 20143, using the actual value of each second candidate input content sample data, it is determined that corresponding each second candidate
The award value of input content.
In actual applications, gradient is calculated for convenience, and actual value is converted into award value.Specifically, calculate first true
The absolute value of difference between real value and 0.5, it is then inverted to obtain corresponding award value.It is appreciated that when actual value more connects
Nearly 0.5, award value is bigger；Conversely, when actual value is further away from 0.5, award value is smaller.
Sub-step 20144, when determining that the completion model is not converged according to the award value, to the completion model
Model parameter is modified to obtain completion model.
In actual applications, it can judge whether completion model restrains according to threshold range.When the average value of award value reaches
During to threshold range, the convergence of completion model, terminate training, obtain final completion model；Otherwise completion model is not converged, modification
Model parameter continues to train, until completion model is restrained.Taken for example, if award value is actual value and the absolute value of 0.5 difference
Inverse, then when the average value of award value reaches more than 100, that is, think that completion model is restrained.It will be appreciated, of course, that threshold value model
Enclose needs to be set according to award value and the relation and practical application scene of actual value, the embodiment of the present invention does not limit it
System.
Specifically, award value is changed by gradient policy and changes model parameter so that be true corresponding to generation next time content
Value is more nearly 0.5, so as to which the simulation input content of generation is truer.As shown in Figure 2 A, award value Reward is changed
The intensive training stage is again introduced into afterwards.
In embodiments of the present invention, the condition generation confrontation network (Conditional being made up of maker and arbiter
Generative Adversarial Network, abbreviation CGAN) model, due to being that text is carried out in embodiments of the present invention
Processing, traditional CGAN generations are discrete results, the bad calculating of gradient, so as to be not easy correction model parameter.And this Shen
Intensive training process is please with the addition of in a processing procedure, goes to be modified model parameter with the award value of intensified learning,
Calculating process is simply, conveniently.
Certainly, the completion model in the embodiment of the present application is also to use other machines model, and the embodiment of the present application is not to it
It is any limitation as.
After completion model construction finishes, then it can enter subsequent use process.
Step 202, if detecting input operation of the user in the page is set, obtain corresponding to the input operation
Input content contextual information；The contextual information includes the contextual information of the setting page.
The step is referred to the detailed description of step 101, will not be repeated here.
Step 203, according to the input content and the contextual information, using default completion model, acquisition meets
The completion content of user view；
The step is referred to the detailed description of step 102, will not be repeated here.
Alternatively, step 203 includes：
Step 2031, the preparatory condition is set according to the input content and the contextual information.
Specifically, input content and contextual information are set into preparatory condition jointly.For example, if input content is " taste
Road ", the comment of vegetable is directed to by context determination current commentary, and has played star comment, so as to which preparatory condition includes：Theme
" vegetable ", emotion " star ", it is related to comment content " taste ".
Step 2032, simulation input content is generated using the generation model of the completion model according to the preparatory condition.
Specifically, the restrictive condition using preparatory condition as generation simulation input content.For example, theme is " vegetable ", feelings
Sense " star ", it is related to comment content " taste ", then simulation input content is to include the star comment of vegetable taste.
Step 2033, judged for the first time using the determining device of the completion model, obtain intermediateness input content.
The step is referred to the detailed description of sub-step 2013, will not be repeated here.
Step 2034, based on the intermediateness input content and the contextual information, generated using the maker
Multiple first candidate input contents.
The step is referred to the detailed description of sub-step 20141, will not be repeated here.
Step 2035, each first candidate input content sample data is judged using the determining device, obtained each
Second candidate's input content sample data and corresponding actual value.
The step is referred to the detailed description of sub-step 20142, will not be repeated here.
Step 2036, sorted forward at least one second candidate input content using actual value, generation and the input
Completion content corresponding to content.
In actual applications, when multiple input contents be present, corresponding completion content is arranged according to actual value descending,
Forward at least one of selected and sorted is recommended.
The completion content generated by the embodiment of the present invention, it is and context-sensitive, improve the recommendation degree of accuracy.Such as Fig. 1 C institutes
Show, commented on for " star ", the probability of negative reviews is larger, so as to which according to user, input content " environment " generates completion content
For " noisy ", rather than " much better ".
Certainly, it is necessary to which explanation, second candidate's input content that the D of completion model is obtained first is a complete sentence
Son, for example be " environment is noisy ", then the word that the embodiment of the present invention then goes selection not input according to the word inputted is recommended,
" noisy " can then be recommended.
Step 204, by the completion commending contents to user.
The step is referred to the detailed description of step 103, will not be repeated here.
Alternatively, step 204 includes：
Step 2041, by the completion content displaying in the specified location for setting the page.
In embodiments of the present invention, by showing that completion content realizes the recommendation of completion content.Wherein, the page is set to use
The current input frame in family, specified location can be the optional position for not influenceing function.For example, in user to certain product or clothes
When business is commented on, by completion content displaying in the lower right corner of comment input frame, or the lower left corner.
It is appreciated that specified location can be the optional position for not influenceing to set page normal function, the embodiment of the present invention
It is not any limitation as.
Alternatively, when it is described set the page as comment input frame when, sub-step 2041 includes sub-step 20411：
Sub-step 20411, by the completion content displaying in the candidate word of the comment input frame adjoining recommends column.
Specifically, because candidate word can exist multiple, so in actual applications, candidate word recommends column to need according to reality
Scene and real needs are sized and form.
When completion content has multiple words, the form of above nextpage is shown, can save space.When user clicks on up and down
During page corresponding button, the completion word of corresponding page is switched to.
Alternatively, step 204 includes：
Sub-step 2042, the candidate in the input method is included using the completion content as the upper screen candidate item of input method
In column.
Such as Fig. 1 C, Fig. 1 F description, first choice that can be using completion content as input method when completion content only has one
Item display.
It should be noted that when having multinomial in completion, can be since preference sequentially toward heel row.
Certainly, in practical application, the embodiment of the present invention can specifically add completion content the dictionary of input, then to it
Higher than the weight order of traditional candidate item, then recommend be can by the completion content can sort traditional candidate item it
Before.
Step 205, when receiving the selection operation for completion content, then the completion content is inputted into the review pages
In the comment input frame in face.
After completion content displaying, selection operation of the user to completion content can be received in corresponding display location,
So as to which completion content is inputted in the comment input frame of review pages.
Alternatively, step 205 includes：
Step 2051, when the selection operation for receiving the completion content recommended the candidate word in column, then by described in
In the completion content input comment input frame.
Wherein, selection operation can be the operation such as click or double-click, and specific implementation form of the present invention to selection operation is not
It is any limitation as.
The embodiment of the present invention can provide user's function of selection in one or more candidate words, and user is selected
Candidate word is shown in input frame as final completion content.
Certainly, in an additional preferred embodiment, step 205 includes：
Sub-step 2052, the selection operation of the upper screen candidate item to being directed to completion content in input method is received, by the benefit
In the full content input comment input frame.As shown in Fig. 1 C, Fig. 1 F, it is identical to shield candidate item on other with input method, Yong Huke
With the upper screen candidate item being shown in by clicking on selection in input method candidate bar, so as to which upper screen candidate item is inputted into comment input frame
In.
In summary, the embodiments of the invention provide one kind to input complementing method, and methods described includes：If detect use
Input operation of the family in the page is set, then obtain input content contextual information corresponding to the input operation；Above and below described
Literary information includes the contextual information of the setting page；According to the input content and the contextual information, using default
Completion model, obtain and meet the completion content of user view；The completion content, which carries, recommends user.Solves first technology
In the candidate word of association is inputted according only to user, it is impossible to embody user view, reduce the enthusiasm of user and liveness and comment
The problem of by quality, can be by contextual informations such as the contextual informations including setting the page, to user in the page is set
Input operation carry out intelligent completion, improve the enthusiasm and liveness and comment quality of user.
Embodiment three
Reference picture 3, it is specific as follows it illustrates a kind of structure chart for inputting complementing device.
Data obtaining module 301, if for detecting input operation of the user in the page is set, obtain described defeated
Enter input content contextual information corresponding to operation；The contextual information includes the contextual information of the setting page.
Completion content obtaining module 302, for according to the input content and the contextual information, utilizing default benefit
Full model, obtain the completion content for meeting user view.
Recommending module 303, for by the completion commending contents to user.
In summary, the embodiments of the invention provide one kind to input complementing device, and described device includes：Acquisition of information mould
Block, if for detecting input operation of the user in the page is set, obtain input content corresponding to the input operation
Contextual information；The contextual information includes the contextual information of the setting page；Completion content obtaining module, for root
According to the input content and the contextual information, using default completion model, the completion content for meeting user view is obtained；
Recommending module, for by the completion commending contents to user.Solves the time for inputting association in first technology according only to user
Select word, it is impossible to embody user view, reduce user enthusiasm and liveness and comment quality the problem of, can by including
The contextual informations such as the contextual information of the page are set, intelligent completion is carried out to input operation of the user in the page is set, carried
The high enthusiasm and liveness and comment quality of user.
Example IV
Reference picture 4, it is specific as follows it illustrates the structure chart of another input complementing device.
Completion model training module 401, for training completion model according to user's sample data；User's sample data
Including：Input content, the contextual information with the affiliated setting page of input content.
Alternatively, in another embodiment of the invention, above-mentioned completion model training module 401 includes：
True content selects submodule, for selecting to meet in the true input of preparatory condition from user's sample data
Hold.
Analog content generates submodule, for the maker using the completion model according to preparatory condition generation pair
The simulation input content answered.
Intermediateness acquisition submodule, in each input for the true input content and simulation input content
Hold, the determining device using the completion model is judged for the first time, obtains corresponding intermediateness input content.
Completion model acquisition submodule, for the maker and described being sentenced again using the intermediateness input content
Other device gives intensive training, obtains the completion model.
Alternatively, in another embodiment of the invention, above-mentioned completion model acquisition submodule, including：
First candidate generation unit, for based on the intermediateness input content and corresponding true input content
Context information, multiple first candidate input contents are generated using the maker.
Second candidate generation unit, for being carried out using the determining device to each first candidate input content sample data
Judge, obtain each second candidate input content sample data and corresponding actual value.
Award value determining unit, for the actual value using each second candidate input content sample data, it is determined that corresponding each
The award value of second candidate's input content.
Continue training unit, for when determining that the completion model is not converged according to the award value, to the completion
The model parameter of model is modified to obtain completion model.
Data obtaining module 402, if for detecting input operation of the user in the page is set, obtain described defeated
Enter input content contextual information corresponding to operation；The contextual information includes the contextual information of the setting page.
Completion content obtaining module 403, for according to the input content and the contextual information, utilizing default benefit
Full model, obtain the completion content for meeting user view；
Alternatively, completion content obtaining module 403 includes：
Preparatory condition sets submodule, for setting the default bar according to the input content and the contextual information
Part.
Analog content generates submodule, for being generated according to the preparatory condition using the generation model of the completion model
Simulation input content.
Medium content acquisition submodule, for being judged for the first time using the determining device of the completion model, obtain middle
State input content.
First alternating content generates submodule, for based on the intermediateness input content and the contextual information,
Multiple first candidate input contents are generated using the maker.
Second alternating content acquisition submodule, for utilizing the determining device to each first candidate input content sample number
According to being judged, each second candidate input content sample data and corresponding actual value are obtained.
Completion content generates submodule, for utilizing the forward at least one second candidate input content of actual value sequence,
Generation completion content corresponding with the input content.
Recommending module 404, for by the completion commending contents to user.
Alternatively, in another embodiment of the invention, above-mentioned recommending module 404, including：
First displaying submodule, for by the completion content displaying it is described setting the page specified location.
Alternatively, in another embodiment of the invention, above-mentioned first displaying submodule, including：
Display unit, for by the completion content displaying it is described comment input frame adjoining candidate word recommend column in.
Alternatively, in another embodiment of the invention, above-mentioned recommending module 404, including：
Second displaying submodule, for including the completion content as the upper screen candidate item of input method in the input
In the candidate bar of method.
Input module 405, for when receiving the selection operation for completion content, then the completion content being inputted into institute
In the comment input frame for stating review pages.
Alternatively, input module 405 includes:
First candidate bar submodule, for as the selection behaviour for receiving the completion content recommended the candidate word in column
Make, then by the completion content input comment input frame.
Alternatively, input module 405 includes:
Second candidate bar submodule, the selection for receiving the upper screen candidate item to being directed to completion content in input method are grasped
Make, by the completion content input comment input frame.
In summary, the embodiments of the invention provide one kind to input complementing device, and described device includes：Acquisition of information mould
Block, if for detecting input operation of the user in the page is set, obtain input content corresponding to the input operation
Contextual information；The contextual information includes the contextual information of the setting page；Completion content obtaining module, for root
According to the input content and the contextual information, using default completion model, the completion content for meeting user view is obtained；
Recommending module, for by the completion commending contents to user.Solves the time for inputting association in first technology according only to user
Select word, it is impossible to embody user view, reduce user enthusiasm and liveness and comment quality the problem of, can by including
The contextual informations such as the contextual information of the page are set, intelligent completion is carried out to input operation of the user in the page is set, carried
The high enthusiasm and liveness and comment quality of user.
The embodiment of the present invention additionally provides a kind of electronic equipment, including：Processor, memory and it is stored in the storage
On device and the computer program that can run on the processor, it is characterised in that real during the computing device described program
The input complementing method of existing previous embodiment.
The embodiment of the present invention additionally provides a kind of readable storage medium storing program for executing, when the instruction in the storage medium is by electronic equipment
Computing device when so that electronic equipment is able to carry out the input complementing method of previous embodiment.
For device embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, it is related
Part illustrates referring to the part of embodiment of the method.
Algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment provided herein.
Various general-purpose systems can also be used together with teaching based on this.As described above, required by constructing this kind of system
Structure be obvious.In addition, the present invention is not also directed to any certain programmed language.It should be understood that it can utilize various
Programming language realizes the content of invention described herein, and the description done above to language-specific is to disclose this hair
Bright preferred forms.
In the specification that this place provides, numerous specific details are set forth.It is to be appreciated, however, that the implementation of the present invention
Example can be put into practice in the case of these no details.In some instances, known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this description.
Similarly, it will be appreciated that in order to simplify the disclosure and help to understand one or more of each inventive aspect,
Above in the description to the exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention：I.e. required guarantor
The application claims of shield features more more than the feature being expressly recited in each claim.It is more precisely, such as following
Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following embodiment are expressly incorporated in the embodiment, wherein each claim is in itself
Separate embodiments all as the present invention.
Those skilled in the art, which are appreciated that, to be carried out adaptively to the module in the equipment in embodiment
Change and they are arranged in one or more equipment different from the embodiment.Can be the module or list in embodiment
Member or component be combined into a module or unit or component, and can be divided into addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit exclude each other, it can use any
Combination is disclosed to all features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so to appoint
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification (including adjoint power
Profit requires, summary and accompanying drawing) disclosed in each feature can be by providing the alternative features of identical, equivalent or similar purpose come generation
Replace.
The all parts embodiment of the present invention can be realized with hardware, or to be run on one or more processor
Software module realize, or realized with combinations thereof.It will be understood by those of skill in the art that it can use in practice
Microprocessor or digital signal processor (DSP) come realize in input completion equipment according to embodiments of the present invention some or
The some or all functions of person's whole part.The present invention is also implemented as perform method as described herein one
Divide either whole equipment or program of device.Such program for realizing the present invention can be stored in computer-readable medium
On, or can have the form of one or more signal.Such signal can be downloaded from internet website and obtained, or
Person provides on carrier signal, or is provided in the form of any other.
It should be noted that the present invention will be described rather than limits the invention for above-described embodiment, and ability
Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between bracket should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" before element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of some different elements and being come by means of properly programmed computer real
It is existing.In if the unit claim of equipment for drying is listed, several in these devices can be by same hardware branch
To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and run after fame
Claim.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, the corresponding process in preceding method embodiment is may be referred to, will not be repeated here.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement made within refreshing and principle etc., should be included in the scope of the protection.
The foregoing is only a specific embodiment of the invention, but protection scope of the present invention is not limited thereto, any
Those familiar with the art the invention discloses technical scope in, change or replacement can be readily occurred in, should all be contained
Cover within protection scope of the present invention.Therefore, protection scope of the present invention should be defined by scope of the claims.
Claims (10)
1. one kind input complementing method, it is characterised in that methods described includes：
If detecting input operation of the user in the page is set, obtain corresponding to the input operation above and below input content
Literary information；The contextual information includes the contextual information of the setting page；
According to the input content and the contextual information, using default completion model, the benefit for meeting user view is obtained
Full content；
By the completion commending contents to user.
2. according to the method for claim 1, it is characterised in that also include：
Completion model is trained according to user's sample data；User's sample data includes：Belonging to input content and input content
Set the contextual information of the page.
3. according to the method for claim 2, it is characterised in that the step that completion model is trained according to user's sample data
Suddenly, including：
Selection meets the true input content of preparatory condition from user's sample data；
Using the completion model maker according to the preparatory condition generation corresponding to simulation input content；
For the true input content and each input content of simulation input content, the determining device of the completion model is utilized
Judged for the first time, obtain corresponding intermediateness input content；
The maker and the arbiter are giveed intensive training again using the intermediateness input content, obtain the benefit
Full model.
4. according to the method for claim 3, it is characterised in that described to utilize the intermediateness input content again to described
Maker and the arbiter give intensive training, the step of obtaining the completion model, including：
Contextual information based on the intermediateness input content and corresponding true input content, is given birth to using the maker
Into multiple first candidate input contents；
Each first candidate input content sample data is judged using the determining device, obtained in each second candidate input
Hold sample data and corresponding actual value；
Using the actual value of each second candidate input content sample data, it is determined that the reward of corresponding each second candidate input content
Value；
When determining that the completion model is not converged according to the award value, the model parameter of the completion model is modified
To obtain completion model.
5. according to the method for claim 4, it is characterised in that described to be believed according to the input content and the context
Breath, using default completion model, the step of meeting the completion content of user view is obtained, including：
The preparatory condition is set according to the input content and the contextual information；
Simulation input content is generated using the generation model of the completion model according to the preparatory condition；
Determining device using the completion model is judged for the first time, obtains intermediateness input content；
Based on the intermediateness input content and the contextual information, it is defeated to generate multiple first candidates using the maker
Enter content；
Each first candidate input content sample data is judged using the determining device, obtained in each second candidate input
Hold sample data and corresponding actual value；
Sorted forward at least one second candidate input content using actual value, generates completion corresponding with the input content
Content.
6. according to the method for claim 1, it is characterised in that when it is described set the page for comment on input frame when, it is described general
The step of completion commending contents are to user, including：
By the completion content displaying in the candidate word of the comment input frame adjoining recommends column.
7. according to the method for claim 1, it is characterised in that it is described by the completion commending contents to user the step of,
Including：
Include the completion content as the upper screen candidate item of input method in the candidate bar of the input method.
8. one kind input complementing device, it is characterised in that described device includes：
Data obtaining module, if for detecting input operation of the user in the page is set, obtain the input operation
Corresponding input content contextual information；The contextual information includes the contextual information of the setting page；
Completion content obtaining module, for according to the input content and the contextual information, using default completion model,
Obtain the completion content for meeting user view；
Recommending module, for by the completion commending contents to user.
9. a kind of electronic equipment, it is characterised in that including：
Processor, memory and the computer program that can be run on the memory and on the processor is stored in, its
It is characterised by, the input completion as described in one or more in claim 1-7 is realized during the computing device described program
Method.
10. a kind of readable storage medium storing program for executing, it is characterised in that when the instruction in the storage medium is held by the processor of electronic equipment
During row so that electronic equipment is able to carry out the input complementing method as described in one or more in claim to a method 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710964623.4A CN107831915A (en) | 2017-10-17 | 2017-10-17 | One kind input complementing method, device, electronic equipment and readable storage medium storing program for executing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710964623.4A CN107831915A (en) | 2017-10-17 | 2017-10-17 | One kind input complementing method, device, electronic equipment and readable storage medium storing program for executing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107831915A true CN107831915A (en) | 2018-03-23 |
Family
ID=61648329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710964623.4A Pending CN107831915A (en) | 2017-10-17 | 2017-10-17 | One kind input complementing method, device, electronic equipment and readable storage medium storing program for executing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107831915A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008331A (en) * | 2019-04-15 | 2019-07-12 | 三角兽（北京）科技有限公司 | Information displaying method, device, electronic equipment and computer readable storage medium |
| CN113031787A (en) * | 2019-12-24 | 2021-06-25 | 北京搜狗科技发展有限公司 | Input method, input device and input device |
| CN113625885A (en) * | 2020-05-08 | 2021-11-09 | 北京搜狗科技发展有限公司 | Input method, input device and input device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130253908A1 (en) * | 2012-03-23 | 2013-09-26 | Google Inc. | Method and System For Predicting Words In A Message |
| CN103365833A (en) * | 2012-03-28 | 2013-10-23 | 百度在线网络技术（北京）有限公司 | Context scene based candidate word input prompt method and system for implementing same |
| CN106527752A (en) * | 2016-09-23 | 2017-03-22 | 百度在线网络技术（北京）有限公司 | Method and device for providing input candidate items |
| CN107168546A (en) * | 2017-03-27 | 2017-09-15 | 上海奔影网络科技有限公司 | Input reminding method and device |
-
2017
- 2017-10-17 CN CN201710964623.4A patent/CN107831915A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130253908A1 (en) * | 2012-03-23 | 2013-09-26 | Google Inc. | Method and System For Predicting Words In A Message |
| CN103365833A (en) * | 2012-03-28 | 2013-10-23 | 百度在线网络技术（北京）有限公司 | Context scene based candidate word input prompt method and system for implementing same |
| CN106527752A (en) * | 2016-09-23 | 2017-03-22 | 百度在线网络技术（北京）有限公司 | Method and device for providing input candidate items |
| CN107168546A (en) * | 2017-03-27 | 2017-09-15 | 上海奔影网络科技有限公司 | Input reminding method and device |
Non-Patent Citations (1)
| Title |
|---|
| YU等: "Sequence Generative Adversarial Nets with Policy Gradient", 《PROCEEDINGS OF THE THIRTY-FIRST AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008331A (en) * | 2019-04-15 | 2019-07-12 | 三角兽（北京）科技有限公司 | Information displaying method, device, electronic equipment and computer readable storage medium |
| CN113031787A (en) * | 2019-12-24 | 2021-06-25 | 北京搜狗科技发展有限公司 | Input method, input device and input device |
| CN113625885A (en) * | 2020-05-08 | 2021-11-09 | 北京搜狗科技发展有限公司 | Input method, input device and input device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111415740B (en) | Method and device for processing inquiry information, storage medium and computer equipment | |
| CN111241311B (en) | Media information recommendation method and device, electronic equipment and storage medium | |
| Brooks et al. | FeatureInsight: Visual support for error-driven feature ideation in text classification | |
| JP6515624B2 (en) | Method of identifying lecture video topics and non-transitory computer readable medium | |
| Holmes et al. | Age of acquisition and typicality effects in three object processing tasks | |
| CN109740048B (en) | Course recommendation method and device | |
| CN110457452A (en) | Rationale for the recommendation generation method, device, electronic equipment and readable storage medium storing program for executing | |
| AU2023202891B2 (en) | Image classification system | |
| Shen et al. | A voice of the customer real-time strategy: An integrated quality function deployment approach | |
| CN117972057A (en) | Dialogue content processing method, equipment, storage medium and product | |
| CN107831915A (en) | One kind input complementing method, device, electronic equipment and readable storage medium storing program for executing | |
| Lipkovich et al. | Multiplicity issues in exploratory subgroup analysis | |
| CN111737558A (en) | Information recommendation method and device and computer readable storage medium | |
| CN112417848A (en) | Corpus generation method and device and computer equipment | |
| US20160350264A1 (en) | Server and method for extracting content for commodity | |
| TW201820172A (en) | System, method and non-transitory computer readable storage medium for conversation analysis | |
| WO2010111340A1 (en) | Teaching system and method | |
| Smith et al. | The use of query auto-completion over the course of search sessions with multifaceted information needs | |
| KR101745874B1 (en) | System and method for a learning course automatic generation | |
| KR20220066882A (en) | Co-information generative adversarial networks for efficient data co-clustering | |
| CN117094387A (en) | Knowledge graph construction method and system based on big data | |
| CN110347916A (en) | Cross-scenario item recommendation method, device, electronic equipment and storage medium | |
| US10296624B2 (en) | Document curation | |
| US20200142963A1 (en) | Apparatus and method for predicting response to an article | |
| JP2001184356A (en) | User taste learning type data base retrieving device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |