CN102968191A - Method for input by using input method and electronic device - Google Patents
Method for input by using input method and electronic device Download PDFInfo
- Publication number
- CN102968191A CN102968191A CN2012104600528A CN201210460052A CN102968191A CN 102968191 A CN102968191 A CN 102968191A CN 2012104600528 A CN2012104600528 A CN 2012104600528A CN 201210460052 A CN201210460052 A CN 201210460052A CN 102968191 A CN102968191 A CN 102968191A
- Authority
- CN
- China
- Prior art keywords
- candidate item
- input
- information
- character
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Machine Translation (AREA)
Abstract
The invention discloses a method for input by using an input method and an electronic device, wherein the method for input by using the input method comprises the following steps of: receiving characters input by using the input method; acquiring candidates corresponding to the characters and prompt information of the candidates; and outputting the candidates and the prompt information of the candidates. In the mode, a user can accurately distinguish each candidate to choose the required candidate so as to finish the input.
Description
Technical field
The present invention relates to input method field, particularly relate to a kind of method and electronic installation that utilizes input method to input.
Background technology
Input method refers to the coding method of employing for various symbols being inputted computing machines or other equipment (such as mobile phone).
Although at present general input method can be supported user key-press input, handwriting input or phonetic entry etc. in input, but the way of output of candidates of input method is also more single, major part is simple show candidate item, then by the user input is finished in the selection of candidate item.
The present patent application people finds in long-term research, there is following problem at least in the way of output of present candidates of input method: some language beginners are difficult to utilize existing input method accurately to select desirable candidate item, or some users are difficult to accurately to offer an explanation the candidate item which is only oneself needs in the face of ambiguity candidate item or polyphone, word the time.
Summary of the invention
The technical matters that the present invention mainly solves provides a kind of method and electronic installation that utilizes input method to input, and can make the user can accurately distinguish each candidate item, and then accurately chooses the candidate item that oneself needs and finish input.
For solving the problems of the technologies described above, the technical scheme that the present invention adopts is: a kind of method of utilizing input method to input is provided, comprises: receive the character that utilizes described input method input; Obtain the information of candidate item corresponding to described character and described candidate item; Information output with described candidate item and described candidate item.
Wherein, after the step of described information output with candidate item and described candidate item, also comprise: detect the instruction whether voice output is arranged; If the instruction of voice output is arranged, then pass through the information of the described candidate item of voice output and described candidate item.
Wherein, the described step of obtaining the information of candidate item corresponding to character and described candidate item comprises: by the conversion dictionary of described input method, obtain candidate item corresponding to described character; According to candidate item corresponding to described character of obtaining, from the large scale network language material, obtain at least one the information as described candidate item in short sentence, the phrase that comprises described candidate item and the example sentence of the explaining described candidate item implication.
Wherein, the described step of obtaining at least one the information as described candidate item in short sentence, the phrase that comprises described candidate item and the example sentence of the explaining described candidate item implication from large-scale corpus comprises: according to the short sentence that comprises described candidate item, phrase and explain that the frequency that the example sentence of described candidate item implication occurs carries out screening first time in the large scale network language material, obtain comprising short sentence, the phrase of described candidate item and explain at least one item in the example sentence of described candidate item implication; Carry out programmed screening according at least one the probability in the n gram language model that builds based on the large scale network language material that screens the described first time in the short sentence that comprises described candidate item, the phrase that obtains and the example sentence of the explaining described candidate item implication, at least one the information as described candidate item that obtains comprising short sentence, the phrase of candidate item and explain the example sentence of described candidate item implication.
Wherein, after the step of described explain information output with candidate item and candidate item, also comprise: receive the input of user selection candidate item, obtain candidate item and the output display of described user selection.
Wherein, after the step of the described candidate item of obtaining user selection and output display, also comprise: according to the input of described user selection candidate item, upgrade the ordering of candidate item in the conversion dictionary of described input method of described user selection.
Wherein, described reception utilizes the step of the character of input method input to comprise: receive the character that utilizes a kind of or any two or more mode in described input method key-press input, handwriting input and the phonetic entry to input.
For solving the problems of the technologies described above, another technical solution used in the present invention is: a kind of electronic installation is provided, comprise receiver module, candidate's acquisition module and output module, wherein: described receiver module is used for receiving the character that utilizes described input method input, and described character is sent to described candidate's acquisition module; Described candidate's acquisition module is used for according to the described character from described receiver module, obtain the information of candidate item corresponding to described character and described candidate item, and the information of the candidate item that described character is corresponding and described candidate item sends to described output module; Described output module is used for and will exports from the described candidate item of described candidate's acquisition module and the information of described candidate item.
Wherein, described device also comprises detection module and voice module, and wherein: described detection module is for detection of the instruction whether voice output is arranged; Described voice module is used for when described detection module detects the instruction of voice output, by the information of the described candidate item of voice output and described candidate item.
Wherein, described candidate's acquisition module comprises the first acquiring unit and second acquisition unit, wherein: described the first acquiring unit is used for receiving the described character from described receiver module, conversion dictionary by described input method, obtain candidate item corresponding to described character, and the described candidate item that will obtain is exported to described second acquisition unit; Described second acquisition unit is used for candidate item corresponding to described character that basis is obtained from described the first acquiring unit, from the large scale network language material, obtain at least one the information as described candidate item in short sentence, the phrase that comprises described candidate item and the example sentence of the explaining described candidate item implication, the information of described candidate item and described candidate item is sent to described output module.
Wherein, described second acquisition unit comprises the first screening subelement and the second screening subelement, wherein: described the first screening subelement is used for candidate item corresponding to described character that basis is obtained from described the first acquiring unit, from the large scale network language material, obtain the short sentence that comprises described candidate item, phrase and explain in the example sentence of described candidate item implication at least one, according to the short sentence that comprises described candidate item, phrase and explain that the frequency that the example sentence of described candidate item implication occurs carries out the screening first time in the large scale network language material obtains comprising the short sentence of described candidate item, phrase and explain in the example sentence of described candidate item implication at least one and export to described the second screening subelement; At least one the probability in the n gram language model that builds based on the large scale network language material that described the second screening subelement is used for the short sentence that comprises described candidate item, the phrase that screening obtains according to the described first time and the example sentence of explaining described candidate item implication carries out programmed screening, at least one the information as described candidate item that obtains comprising short sentence, the phrase of candidate item and explain the example sentence of described candidate item implication, and the information of described candidate item and described candidate item sent to described output module.
Wherein, also comprise the candidate display module, be used for to receive the user and select the input of described candidate item according to the information of the described candidate item of described output module output and described candidate item, obtain candidate item and the output display of described user selection.
Wherein, described device also comprises candidate's update module, is used for the input according to user selection candidate item described in the described candidate display module, upgrades the ordering of candidate item in the conversion dictionary of described input method of described user selection.
Wherein, described receiver module specifically be used for to receive the character that a kind of or any two or more mode of utilizing described input method key-press input, handwriting input and phonetic entry is inputted, and described character is sent to described candidate's acquisition module.
The invention has the beneficial effects as follows: the situation that is different from prior art, the method that the present invention utilizes input method to input, utilize the information of candidate item, can further explain candidate item, so that the user accurately distinguishes each candidate item, when the user faces the input of the corresponding a plurality of Chinese words of same assumed name in homophony words in the spelling input method or the Japanese inputting method, can be good at assisted user and accurately select candidate item and finish input.In addition, for the beginner of a language, more can better grasp this language by the study to the information (such as example sentence etc.) of candidate item.
Description of drawings
Fig. 1 is the process flow diagram that the present invention utilizes method one embodiment that input method inputs;
Fig. 2 is the process flow diagram that the present invention utilizes another embodiment of method that input method inputs;
Fig. 3 be the present invention utilize input method to input method one embodiment in obtain the process flow diagram of the information of candidate item corresponding to character and candidate item;
Fig. 4 be the present invention utilize input method to input method one embodiment in from large-scale corpus, obtain the short sentence that comprises candidate item, phrase and explain at least one the process flow diagram as the information of candidate item in the example sentence of candidate item implication;
Fig. 5 is the structural representation of electronic installation one embodiment of the present invention;
Fig. 6 is the structural representation of candidate's acquisition module in electronic installation one embodiment of the present invention;
Fig. 7 is the structural representation of second acquisition unit in electronic installation one embodiment of the present invention;
Fig. 8 is the structural representation of another embodiment of electronic installation of the present invention.
Embodiment
See also Fig. 1, method one embodiment that the present invention utilizes input method to input comprises:
Step S101: receive the character that utilizes the input method input;
In the embodiment of the present invention, input refer to by press key on the lower keyboard, hand-written or voice mode send character to device.Send character " f " such as the letter " f " of pressing on the lower keyboard to device, again such as sending character " Off " by drawing " Off " at touch-screen to device, again or by starting the phonetic function of electronic installation, says " university " facing to device, to device transmission phonetic characters " university ".Character refers to the letter, numeral, word and the symbol that use in the computing machine, such as 1,2,3, A, B, C,! ,? etc..
Receive the character that the user utilizes the input method input, the character of input can be the character of inputting by a kind of or any two or more mode in key-press input, handwriting input and the phonetic entry here.
Step S102: the information of obtaining candidate item corresponding to character and candidate item;
Input to the user is resolved, and obtains the character of user's input and the character of inputting according to the user, obtains the information of candidate item corresponding to character and candidate item.The information that is used for distinguishing each candidate item that provides in order to make the user can accurately select the candidate item that oneself needs from a plurality of candidate item that acquire is provided the information of the candidate item here, can be one or any information more than two of the short sentence, phrase and the candidate item implication that comprise candidate item.
Character such as user's input is " haoma ", the candidate item of obtaining according to character can be " number ", " OK ", " fine horse " ..., and the information of the candidate item of the character that obtains can be: such as can being " telephone number ", " phone number ", " QQ number " etc. with information corresponding to " number "; The information corresponding with " OK " can be " how do you do ", " OK recently " etc.; Information with " fine horse " correspondence can be " fine horse ", " fine horse is known the way " etc.
Step S103: with the information output of candidate item and candidate item;
With the information output of the candidate item obtained and candidate item and be presented on the screen, so that the user selects the candidate item of own needs according to the information of the candidate item of output and corresponding candidate item.
The user can select candidate item by a kind of or any two or more mode in key-press input, handwriting input, the phonetic entry.Device end receives the input of user selection candidate item, obtains corresponding candidate item and the output display of user selection by identification.
In actual application, can also according to the input of the candidate item of user selection, upgrade the ordering of candidate item in the conversion dictionary of input method of user selection.The conversion dictionary of the input method here comprises the assumed name of the phonetic of spelling input method-Chinese character conversion dictionary, Japanese inputting method-Chinese character conversion dictionary or voice-Chinese character conversion dictionary etc., does not give unnecessary details one by one at this.
Such as user's input " haoma ", in continuous 5 inputs, have and selected candidate item " number " 4 times, once selected candidate item " OK ", first in the candidate item that at this moment can " haoma " in phonetic-Chinese character conversion dictionary is corresponding be updated to " number ".Certain this renewal has real-time, has all selected " OK " such as the user in follow-up 5 inputs " haoma ", at this moment first of corresponding candidate can be updated to " OK ".Equally, in Japanese inputting method, if user 4 times input assumed name " い い ん " has been selected candidate " committee member " 3 times, first of candidate item that then can " い い ん " in assumed name-Chinese character conversion dictionary is corresponding is updated to " committee member ".When the user uses phonetic entry, also can be by such mode, progressively reducing under the prerequisite of user interactions, improving the accuracy rate that to user speech identification obtains the needed candidate item of user (at first candidate item and user input the matching rate of expecting the candidate item that obtains namely to improve input method).
Elaboration by above-mentioned embodiment, be appreciated that, the method that the present invention utilizes input method to input, utilize the information of candidate item, can further explain candidate item, so that the user can accurately distinguish each candidate item, when the user faces the input of the corresponding a plurality of Chinese words of same assumed name in homophony words in the spelling input method or the Japanese inputting method, can be good at assisted user and accurately select candidate item and finish input.In addition, for the beginner of a language, more can better grasp this language by the study to the information (such as example sentence etc.) of candidate item.
In actual application, the environment that it is poor that the user may face light condition is difficult to clearly see candidate item and the candidate item information of demonstration, and perhaps eye disease patient is difficult to utilize existing input method to finish input such as the blind person.Therefore, the invention provides another embodiment of method that utilizes input method to input, can make the general user at light bad or eye disease patient also can accurately select candidate item and finish input, see also Fig. 2, present embodiment may further comprise the steps:
Step S201: receive the character that utilizes the input method input;
Receive the character that the user utilizes the input method input, the character of input can be the character of inputting by a kind of or any two or more mode in key-press input, handwriting input and the phonetic entry here.
Step S202: the information of obtaining candidate item corresponding to character and candidate item;
Input to the user is resolved, and obtains the character of user's input and the character of inputting according to the user, obtains the information of candidate item corresponding to character and candidate item.The information that is used for distinguishing each candidate item that provides in order to make the user can accurately select the candidate item that oneself needs from a plurality of candidate item that acquire is provided the information of the candidate item here, can be one or any information more than two of the short sentence, phrase and the candidate item implication that comprise candidate item.
Step S203: with the information output of candidate item and candidate item;
The information of candidate item and candidate item is exported and is presented on the screen, for user selection.
Step S204: detect the instruction whether voice output is arranged;
When the user faces the bad environment of light or eye disease patient and need to utilize input method to input, can trigger voice output by corresponding button, induction or other mode.Device detects the instruction whether voice output is arranged, when using the instruction of voice output, enter step S205, if do not detect the instruction of voice output, represent that then the active user does not need to come the assisted Selection candidate item by voice output, directly keep current output mode to wait the user selection candidate item.
Step S205: by the information of voice output candidate item and candidate item;
When detecting the instruction of voice output, export to the user by candidate item and information that the mode of voice will be exported and show, allow the user can accurately select the candidate item that oneself needs.
By with upper type, to the information of candidates of input method and candidate item with the formal output of voice to the user, even the user in the poor situation of sighting condition or eye disease patient also can know and select candidate item accurately and finish input.
See also Fig. 3, the information of obtaining candidate item corresponding to character and candidate item in another embodiment of method that the present invention utilizes input method to input comprises following substep:
Substep S301: according to the conversion dictionary of input method, obtain candidate item corresponding to character;
Character according to user's input utilizes the corresponding conversion dictionary of input method to obtain candidate item corresponding to character.Such as user's input " yinhang ", from phonetic-Chinese character conversion dictionary, obtain corresponding candidate item " bank ", " pilotage ", " Yin Hang " etc.; Or user input " い い ん ", from assumed name-Chinese character conversion dictionary, obtain corresponding candidate item “ Wei STAFF ", " hospital ", “ Yi STAFF " etc.
Substep S302: the candidate item corresponding according to the character that obtains, from the large scale network language material, obtain at least one the information as candidate item in short sentence, the phrase that comprises candidate item and the example sentence of the explaining the candidate item implication;
The candidate item corresponding according to the character that obtains obtained information corresponding to each candidate item.The information of these candidate item can be the short sentence that comprises candidate item, phrase and the example sentence of explaining the candidate item implication etc. one or multinomial.Such as the assumed name " い い ん " for user input, candidate item may comprise “ Wei STAFF ", " hospital " etc., can distinguish this two candidates by sub " can entrust " (the meeting committee member) of short sentence and " the capable く of the へ of hospital " (going to hospital).
See also Fig. 4, at least one the information as candidate item of obtaining from large-scale corpus in another embodiment of method that the present invention utilizes input method to input in short sentence, the phrase that comprises candidate item and the example sentence of the explaining the candidate item implication comprises following substep:
Substep S401: according to the short sentence that comprises candidate item, phrase and explain that the frequency that the example sentence of candidate item implication occurs carries out screening first time in the large scale network language material, obtain comprising short sentence, the phrase of candidate item and explain at least one item in the example sentence of candidate item implication;
In order to use as much as possible minimum information to distinguish each candidate item, can screen by the information to candidate item, to obtain more appropriate candidate item information.At first, all that obtain are comprised short sentence, the phrase of candidate item and explain that the example sentence of candidate item implication just sorts according to the frequency, only get the short sentence that comprises candidate item, the phrase that wherein a part of frequency is high and the example sentence of explaining the candidate item implication.The employing frequency is filtered, can be as far as possible those " commonly use " and sentence or phrase etc. return to the user as prompting because the number of times that occurs in webpage is more, then we think, this short sentence has more been used by most people.
Such as for same candidate item, the information that can be used as candidate item of obtaining four of A, B, C, D are arranged, if the frequency that A, B, C, D occur in the large scale network language material just is respectively A〉B〉C〉D, can only stay A, B, C or A, B or A.
Substep S402: carry out programmed screening according at least one the probability in the n gram language model that builds based on the large scale network language material that screens for the first time in the short sentence that comprises candidate item, the phrase that obtains and the example sentence of the explaining the candidate item implication, at least one the information as candidate item that obtains comprising short sentence, the phrase of candidate item and explain the example sentence of candidate item implication;
The probability that current word occurs in the n gram language model (n-gram language model) only has relation with n-1 the word on its left side.When the n-gram language model is used Chinese web page, obtain Chinese n gram language model; When the n-gram language model is used English webpage, obtain English n gram language model.For example when the n value was 2, the probability of the appearance of current word only had relation with its previous word.For example for sentence:
S=Zhang San chairman of the board has delivered the speech of four preferential important indications.
Under 2 gram language model, the probability of this sentence (weighing the tolerance of the correctness of this sentence) is:
P (S)=P (Zhang San |＜s 〉) P (chairman of the board | Zhang San) P (deliver | the chairman of the board) P (| deliver) P (four |) P (preferential | four) P (important | preferential) P (indication | important) P (| indication) P (speech |) P (.| speech) P (＜/s〉|.)
Here＜s〉and＜/s 〉, be the word of two manual construction, represented respectively beginning and the ending of sentence.(its objective is judgement " Zhang San " as the probability of sentence entry word, and "." fullstop is as the probability of sentence suffixed word)
If under 3 gram language model, the probability of this sentence is:
P (S)=P (Zhang San |＜s 〉) P (chairman of the board |＜s 〉, Zhang San) P (deliver | Zhang San, the chairman of the board) P (| the chairman of the board, deliver) P (four | deliver) P (preferential |, three) and P (important | four, preferentially) P (indication | preferential, important) P (| important, indication) P (speech | indication) P (.|, speech) P (＜/s〉| speech.)
Here, the computing method of a probability are in 2 meta-models:
P (chairman of the board | Zhang San)=count (Zhang San chairman of the board)/count (Zhang San)
Molecule is the frequency that " Zhang San chairman of the board " occurs in corpus (for example large scale network language material); Denominator is the frequency that " Zhang San " occurs in corpus.
Correspondingly, the computing formula of a probability is in 3 meta-models:
P (deliver | Zhang San, chairman of the board)=count (Zhang San chairman of the board delivers)/count (Zhang San chairman of the board)
The molecule here is the frequency that " Zhang San chairman of the board delivers " occurs in corpus, and denominator is the frequency that " Zhang San chairman of the board " occurs in corpus.
According to the probability sorting in the n gram language model, filter out short sentence, phrase or the example sentence etc. of corresponding information as candidate item for the remaining short sentence that comprises candidate item of for the first time screening, phrase and the example sentence of explaining the candidate item implication.Adopting probability to carry out the second time and filter, is the problem of considering " discrimination " of candidate's sentence.Because probability is the ratio of two frequencys, so even the low sentence of the frequency may because it has comprised regular collocation, have the character (probability that namely occurs is very high) of " high cohesion " on the contrary in the n gram language model.
Such as being left A, B, C through for the first time screening, wherein probability is respectively A:0.4, B:0.2, C:0.1, can be with A, B as exporting at last the information of user as the candidate item of prompting, if certainly only keep a candidate item information, then only with A as the information of exporting at last the user.
Mode by above-mentioned twice screening, obtain the relatively high and probability of frequency of usage also the not low short sentence that comprises candidate item, phrase and explain one of example sentence of the candidate item implication or any more than two the information as candidate item return to the user, reached the purpose of " commonly used, high discrimination ".
Such as hypothesis user's input " yiyuan ", its corresponding Chinese character candidate has: hospital, wish, monobasic etc.For " monobasic " this candidate, corresponding information may be included in short sentence such as " yuans; function of a single variable; quadratic equation with one unknown ", and probably in our corpus " yuan " be not that frequency of occurrence is best, but because this is " compound word ", its probability on language model is very high.According to the strategy of our twice screening, namely " first frequency posterior probability " filters and ordering, and we can obtain " yuan " and return to the user as first-selection.Because this is widely known by the people, impressive.
See also Fig. 5, electronic installation one embodiment of the present invention comprises receiver module 11, candidate's acquisition module 12 and output module 13, wherein:
In the embodiment of the present invention, input refer to by press key on the lower keyboard, hand-written or voice mode send character to device.Send character " f " such as the letter " f " of pressing on the lower keyboard to device, again such as sending character " Off " by drawing " Off " at touch-screen to device, again or by starting the phonetic function of electronic installation, says " university " facing to device, to device transmission phonetic characters " university ".Character refers to the letter, numeral, word and the symbol that use in the computing machine, such as 1,2,3, A, B, C,! ,? etc..
Candidate's acquisition module 12 is used for obtain the information of candidate item corresponding to character and candidate item, and the information of the candidate item that character is corresponding and candidate item sending to output module 13 according to the character from receiver module 11;
Candidate's acquisition module 12 obtains the information of candidate item corresponding to character and candidate item according to the character from user's input of receiver module 11, and the candidate item obtained and the information of candidate item are sent to output module 13.
Input to the user is resolved, and obtains the character of user's input and the character of inputting according to the user, obtains the information of candidate item corresponding to character and candidate item.The information that is used for distinguishing each candidate item that provides in order to make the user can accurately select the candidate item that oneself needs from a plurality of candidate item that acquire is provided the information of the candidate item here, can be one or any information more than two of the short sentence, phrase and the candidate item implication that comprise candidate item.
Character such as user's input is " haoma ", the candidate item of obtaining according to character can be " number ", " OK ", " fine horse " ..., and the information of the candidate item of the character that obtains can be: such as can being " telephone number ", " phone number ", " QQ number " etc. with information corresponding to " number "; The information corresponding with " OK " can be " how do you do ", " OK recently " etc.; Information with " fine horse " correspondence can be " fine horse ", " fine horse is known the way " etc.
The information output that the candidate item that output module 13 is corresponding with character and corresponding candidate item are corresponding also is shown to the user.
See also Fig. 6, in another embodiment of electronic installation of the present invention, candidate's acquisition module comprises the first acquiring unit 111 and second acquisition unit 112, wherein:
The first acquiring unit 111 is used for receiving the described character from receiver module, by the conversion dictionary of input method, obtains candidate item corresponding to character, and the candidate item of obtaining is exported to second acquisition unit 112;
The first acquiring unit 111 can be by input method assumed name-Chinese character conversion dictionary, phonetic-Chinese character conversion dictionary etc. obtain candidate item corresponding to character, and the candidate item of obtaining exported to second acquisition unit 112.
Character according to user's input utilizes the corresponding conversion dictionary of input method to obtain candidate item corresponding to character.Such as user's input " yinhang ", from phonetic-Chinese character conversion dictionary, obtain corresponding candidate item " bank ", " pilotage ", " Yin Hang " etc.; Or user input " い い ん ", from assumed name-Chinese character conversion dictionary, obtain corresponding candidate item “ Wei STAFF ", " hospital ", “ Yi STAFF " etc.
The candidate item corresponding according to the character that obtains obtained information corresponding to each candidate item.The information of these candidate item can be the short sentence that comprises candidate item, phrase and the example sentence of explaining the candidate item implication etc. one or multinomial.Such as the assumed name " い い ん " for user input, candidate item may comprise “ Wei STAFF ", " hospital " etc., can distinguish this two candidates by sub " can entrust " (the meeting committee member) of short sentence and " the capable く of the へ of hospital " (going to hospital).
See also Fig. 7, in another embodiment of electronic installation of the present invention, second acquisition unit comprises the first screening subelement 1111 and the second screening subelement 1112, wherein:
The first screening subelement 1111 is used for according to candidate item corresponding to character of obtaining from the first acquiring unit, from the large scale network language material, obtain at least one in short sentence, the phrase that comprises candidate item and the example sentence of the explaining the candidate item implication, according to the short sentence that comprises candidate item, phrase and explain that the frequency that the example sentence of candidate item implication occurs carries out screening first time in the large scale network language material, obtain comprising short sentence, the phrase of candidate item and explain in the example sentence of candidate item implication at least one and export to second and screen subelement 1112;
In the short sentence that comprises candidate item as the information of candidate item, the phrase that the first 1111 pairs on subelement of screening gets access to and the example sentence of explaining candidate item at least one carries out screening first time according to the frequency, at least one item in the short sentence that comprises candidate item, the phrase that obtain through for the first time screening and the example sentence of explaining candidate item exported to second screen subelement 1112.
In order to use as much as possible minimum information to distinguish each candidate item, can screen by the information to candidate item, to obtain more appropriate candidate item information.At first, all that obtain are comprised short sentence, the phrase of candidate item and explain that the example sentence of candidate item implication just sorts according to the frequency, only get the short sentence that comprises candidate item, the phrase that wherein a part of frequency is high and the example sentence of explaining the candidate item implication.The employing frequency is filtered, can be as far as possible those " commonly use " and sentence or phrase etc. return to the user as prompting because the number of times that occurs in webpage is more, then we think, this short sentence has more been used by most people.
Such as for same candidate item, the information that can be used as candidate item of obtaining four of A, B, C, D are arranged, if the frequency that A, B, C, D occur in the large scale network language material just is respectively A〉B〉C〉D, can only stay A, B, C or A, B or A.
At least one the probability in the n gram language model that builds based on the large scale network language material that the second screening subelement 1112 is used for the short sentence that comprises candidate item, the phrase that obtains according to for the first time screening and the example sentence of explaining the candidate item implication carries out programmed screening, at least one the information as candidate item that obtains comprising short sentence, the phrase of candidate item and explain the example sentence of described candidate item implication, and the information of candidate item and candidate item sent to output module.
At least one in the short sentences that comprise candidate item, the phrases that the second for the first time screening of 1112 pairs on subelement of screening obtains and the example sentence of explaining candidate item is carried out programmed screening according to probability, with through the short sentence that comprises candidate item, phrase that programmed screening obtains and explain that the candidate item that in the example sentence of candidate item at least one and the first acquiring unit obtain sends to output module together.
According to the probability sorting in the n gram language model, filter out short sentence, phrase or the example sentence etc. of corresponding information as candidate item for the remaining short sentence that comprises candidate item of for the first time screening, phrase and the example sentence of explaining the candidate item implication.Adopting probability to carry out the second time and filter, is the problem of considering " discrimination " of candidate's sentence.Because probability is the ratio of two frequencys, so even the low sentence of the frequency may because it has comprised regular collocation, have the character (probability that namely occurs is very high) of " high cohesion " on the contrary in the n gram language model.
Such as being left A, B, C through for the first time screening, wherein probability is respectively A:0.4, B:0.2, C:0.1, can be with A, B as exporting at last the information of user as the candidate item of prompting, if certainly only keep a candidate item information, then only with A as the information of exporting at last the user.
Mode by above-mentioned twice screening, obtain the relatively high and probability of frequency of usage also the not low short sentence that comprises candidate item, phrase and explain one of example sentence of the candidate item implication or any more than two the information as candidate item return to the user, reached the purpose of " commonly used, high discrimination ".
Such as hypothesis user's input " yiyuan ", its corresponding Chinese character candidate has: hospital, wish, monobasic etc.For " monobasic " this candidate, corresponding information may be included in short sentence such as " yuans; function of a single variable; quadratic equation with one unknown ", and probably in our corpus " yuan " be not that frequency of occurrence is best, but because this is " compound word ", its probability on language model is very high.According to the strategy of our twice screening, namely " first frequency posterior probability " filters and ordering, and we can obtain " yuan " and return to the user as first-selection.Because this is widely known by the people, impressive.
See also Fig. 8, another embodiment of device that the present invention utilizes input method to input comprises receiver module 21, candidate's acquisition module 22, output module 23, detection module 26, voice module 27, candidate display module 24 and candidate's update module 25, wherein:
Candidate's acquisition module 22 is used for obtain the information of candidate item corresponding to character and candidate item, and the information of the candidate item that character is corresponding and candidate item sending to output module 23 according to the character from receiver module 21;
Candidate's update module 25 is used for the input according to candidate display module 24 user selection candidate item, upgrades the ordering of candidate item in the conversion dictionary of input method of user selection;
Candidate's update module 25 is according to the input of user selection candidate item, the ordering of the candidate item of real-time update user selection in input method conversion dictionary.
The user can select candidate item by a kind of or any two or more mode in key-press input, handwriting input, the phonetic entry.Device end receives the input of user selection candidate item, obtains corresponding candidate item and the output display of user selection by identification.
In actual application, can also according to the input of the candidate item of user selection, upgrade the ordering of candidate item in the conversion dictionary of input method of user selection.The conversion dictionary of the input method here comprises the assumed name of the phonetic of spelling input method-Chinese character conversion dictionary, Japanese inputting method-Chinese character conversion dictionary or voice-Chinese character conversion dictionary etc., does not give unnecessary details one by one at this.
Such as user's input " haoma ", in continuous 5 inputs, have and selected candidate item " number " 4 times, once selected candidate item " OK ", first in the candidate item that at this moment can " haoma " in phonetic-Chinese character conversion dictionary is corresponding be updated to " number ".Certain this renewal has real-time, has all selected " OK " such as the user in follow-up 5 inputs " haoma ", at this moment first of corresponding candidate can be updated to " OK ".Equally, in Japanese inputting method, if user 4 times input assumed name " い い ん " has been selected candidate " committee member " 3 times, first of candidate item that then can " い い ん " in assumed name-Chinese character conversion dictionary is corresponding is updated to " committee member ".When the user uses phonetic entry, also can be by such mode, progressively reducing under the prerequisite of user interactions, improving the accuracy rate that to user speech identification obtains the needed candidate item of user (at first candidate item and user input the matching rate of expecting the candidate item that obtains namely to improve input method).
When detection module 26 detected the instruction of voice output, candidate item and information that voice module 27 will be exported and show by the mode of voice were exported to the user, allowed the user can accurately select the candidate item that oneself needs.
Elaboration by above-mentioned embodiment, be appreciated that, the method that the present invention utilizes input method to input, by to the information of candidates of input method and candidate item with the formal output of voice to the user, even the user in the poor situation of sighting condition or eye disease patient also can know and select candidate item accurately and finish input.And utilize the information of candidate item, can further explain candidate item, when the user faces the input of the corresponding a plurality of Chinese words of same assumed name in homophony words in the phonetic output method or the Japanese inputting method, can be good at assisted user and accurately select candidate item and finish input.In addition, for the beginner of a language, more can better grasp this language by the study to the information (such as example sentence etc.) of candidate item.
On the other hand, the interactive voice result is with the final identification that improves the user speech input of the form of Active Learning, and study and accumulation by user's input behavior is selected, constantly update the candidate item ordering of candidate item in input method conversion dictionary, gradually reduce successively user's prompting, finally be expected under the prerequisite of 0 prompting, realize the input of pin-point accuracy.
In several embodiments provided by the present invention, should be understood that disclosed apparatus and method can realize by another way.For example, device embodiments described above only is schematic, for example, the division of described module, only be that a kind of logic function is divided, during actual the realization other dividing mode can be arranged, for example a plurality of modules or assembly can in conjunction with or can be integrated into another system, or some features can ignore, or do not carry out.Another point, the shown or coupling each other discussed or direct-coupling or communication connection can be by some interfaces, indirect coupling or the communication connection of device or unit can be electrically, machinery or other form.
Described functional module as the separating component explanation can or can not be physically to separate also, the parts that show as the unit can be or can not be physical locations also, namely can be positioned at a place, perhaps also can be distributed on a plurality of network element.Can select according to the actual needs wherein some or all of unit to realize the present invention program's purpose.
In addition, each functional module in each embodiment of the present invention can be integrated in the processing unit, also can be that the independent physics of each functional module exists, and also can two or more functional modules be integrated in the unit.Above-mentioned integrated unit both can adopt the form of hardware to realize, also can adopt the form of SFU software functional unit to realize.
The above only is embodiments of the present invention; be not so limit claim of the present invention; every equivalent structure or equivalent flow process conversion that utilizes instructions of the present invention and accompanying drawing content to do; or directly or indirectly be used in other relevant technical fields, all in like manner be included in the scope of patent protection of the present invention.
Claims (14)
1. a method of utilizing input method to input is characterized in that, comprising:
Reception utilizes the character of described input method input;
Obtain the information of candidate item corresponding to described character and described candidate item;
Information output with described candidate item and described candidate item.
2. method according to claim 1 is characterized in that, after the step of described information output with candidate item and described candidate item, also comprises:
Detect the instruction whether voice output is arranged;
If the instruction of voice output is arranged, then pass through the information of the described candidate item of voice output and described candidate item.
3. method according to claim 1 is characterized in that, the described step of obtaining the information of candidate item corresponding to character and described candidate item comprises:
By the conversion dictionary of described input method, obtain candidate item corresponding to described character;
According to candidate item corresponding to described character of obtaining, from the large scale network language material, obtain at least one the information as described candidate item in short sentence, the phrase that comprises described candidate item and the example sentence of the explaining described candidate item implication.
4. method according to claim 3, it is characterized in that described at least one the step as the information of described candidate item of obtaining in short sentence, the phrase that comprises described candidate item and the example sentence of the explaining described candidate item implication comprises from large-scale corpus:
According to the short sentence that comprises described candidate item, phrase and explain that the frequency that the example sentence of described candidate item implication occurs carries out the screening first time in the large scale network language material, obtain comprising short sentence, the phrase of described candidate item and explain in the example sentence of described candidate item implication at least one;
Carry out programmed screening according at least one the probability in the n gram language model that builds based on the large scale network language material that screens the described first time in the short sentence that comprises described candidate item, the phrase that obtains and the example sentence of the explaining described candidate item implication, at least one the information as described candidate item that obtains comprising short sentence, the phrase of candidate item and explain the example sentence of described candidate item implication.
5. method according to claim 1 is characterized in that, after the step of described explain information output with candidate item and candidate item, also comprises: receive the input of user selection candidate item, obtain candidate item and the output display of described user selection.
6. method according to claim 5 is characterized in that, after the step of the described candidate item of obtaining user selection and output display, also comprises:
According to the input of described user selection candidate item, upgrade the ordering of candidate item in the conversion dictionary of described input method of described user selection.
7. method according to claim 1, it is characterized in that described reception utilizes the step of the character of input method input to comprise: receive the character that utilizes a kind of or any two or more mode in described input method key-press input, handwriting input and the phonetic entry to input.
8. an electronic installation is characterized in that, comprises receiver module, candidate's acquisition module and output module, wherein:
Described receiver module is used for receiving the character that utilizes described input method input, and described character is sent to described candidate's acquisition module;
Described candidate's acquisition module is used for according to the described character from described receiver module, obtain the information of candidate item corresponding to described character and described candidate item, and the information of the candidate item that described character is corresponding and described candidate item sends to described output module;
Described output module is used for and will exports from the described candidate item of described candidate's acquisition module and the information of described candidate item.
9. device according to claim 8 is characterized in that, described device also comprises detection module and voice module, wherein:
Described detection module is for detection of the instruction whether voice output is arranged;
Described voice module is used for when described detection module detects the instruction of voice output, by the information of the described candidate item of voice output and described candidate item.
10. device according to claim 8 is characterized in that, described candidate's acquisition module comprises the first acquiring unit and second acquisition unit, wherein:
Described the first acquiring unit is used for receiving the described character from described receiver module, by the conversion dictionary of described input method, obtain candidate item corresponding to described character, and the described candidate item that will obtain is exported to described second acquisition unit;
Described second acquisition unit is used for candidate item corresponding to described character that basis is obtained from described the first acquiring unit, from the large scale network language material, obtain at least one the information as described candidate item in short sentence, the phrase that comprises described candidate item and the example sentence of the explaining described candidate item implication, the information of described candidate item and described candidate item is sent to described output module.
11. device according to claim 9 is characterized in that, described second acquisition unit comprises the first screening subelement and the second screening subelement, wherein:
Described the first screening subelement is used for candidate item corresponding to described character that basis is obtained from described the first acquiring unit, from the large scale network language material, obtain the short sentence that comprises described candidate item, phrase and explain in the example sentence of described candidate item implication at least one, according to the short sentence that comprises described candidate item, phrase and explain that the frequency that the example sentence of described candidate item implication occurs carries out the screening first time in the large scale network language material obtains comprising the short sentence of described candidate item, phrase and explain in the example sentence of described candidate item implication at least one and export to described the second screening subelement;
At least one the probability in the n gram language model that builds based on the large scale network language material that described the second screening subelement is used for the short sentence that comprises described candidate item, the phrase that screening obtains according to the described first time and the example sentence of explaining described candidate item implication carries out programmed screening, at least one the information as described candidate item that obtains comprising short sentence, the phrase of candidate item and explain the example sentence of described candidate item implication, and the information of described candidate item and described candidate item sent to described output module.
12. device according to claim 8, it is characterized in that, also comprise the candidate display module, be used for to receive the user and select the input of described candidate item according to the information of the described candidate item of described output module output and described candidate item, obtain candidate item and the output display of described user selection.
13. device according to claim 12, it is characterized in that, described device also comprises candidate's update module, is used for the input according to user selection candidate item described in the described candidate display module, upgrades the ordering of candidate item in the conversion dictionary of described input method of described user selection.
14. device according to claim 8, it is characterized in that, described receiver module specifically be used for to receive the character that a kind of or any two or more mode of utilizing described input method key-press input, handwriting input and phonetic entry is inputted, and described character is sent to described candidate's acquisition module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012104600528A CN102968191A (en) | 2012-11-15 | 2012-11-15 | Method for input by using input method and electronic device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012104600528A CN102968191A (en) | 2012-11-15 | 2012-11-15 | Method for input by using input method and electronic device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102968191A true CN102968191A (en) | 2013-03-13 |
Family
ID=47798370
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012104600528A Pending CN102968191A (en) | 2012-11-15 | 2012-11-15 | Method for input by using input method and electronic device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102968191A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106125955A (en) * | 2016-06-23 | 2016-11-16 | 百度在线网络技术（北京）有限公司 | A kind of method and apparatus that hot word is provided in applying in input method |
| CN107451036A (en) * | 2017-06-19 | 2017-12-08 | 阿里巴巴集团控股有限公司 | Input reminding method, device and equipment |
| CN115407882A (en) * | 2022-07-13 | 2022-11-29 | 穆运洋 | Visualization-based big data analysis and arrangement system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102063451A (en) * | 2010-04-16 | 2011-05-18 | 百度在线网络技术（北京）有限公司 | Method and equipment for inputting characters by user and providing relevant search information |
| CN102446061A (en) * | 2010-10-06 | 2012-05-09 | 富士通株式会社 | Information terminal apparatus, and character input method |
| CN102541276A (en) * | 2010-12-15 | 2012-07-04 | 富泰华工业（深圳）有限公司 | Input method system with word-meaning association function and input method |
-
2012
- 2012-11-15 CN CN2012104600528A patent/CN102968191A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102063451A (en) * | 2010-04-16 | 2011-05-18 | 百度在线网络技术（北京）有限公司 | Method and equipment for inputting characters by user and providing relevant search information |
| CN102446061A (en) * | 2010-10-06 | 2012-05-09 | 富士通株式会社 | Information terminal apparatus, and character input method |
| CN102541276A (en) * | 2010-12-15 | 2012-07-04 | 富泰华工业（深圳）有限公司 | Input method system with word-meaning association function and input method |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106125955A (en) * | 2016-06-23 | 2016-11-16 | 百度在线网络技术（北京）有限公司 | A kind of method and apparatus that hot word is provided in applying in input method |
| CN106125955B (en) * | 2016-06-23 | 2019-05-07 | 百度在线网络技术（北京）有限公司 | A kind of method and apparatus for the offer hot word in input method is applied |
| CN107451036A (en) * | 2017-06-19 | 2017-12-08 | 阿里巴巴集团控股有限公司 | Input reminding method, device and equipment |
| CN115407882A (en) * | 2022-07-13 | 2022-11-29 | 穆运洋 | Visualization-based big data analysis and arrangement system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107291783B (en) | Semantic matching method and intelligent equipment | |
| CN101350004B (en) | Method for forming personalized error correcting model and input method system of personalized error correcting | |
| CN1918578B (en) | Handwriting and voice input with automatic correction | |
| CN102520874B (en) | Pinyin input method based on touch screen and device | |
| CN105164616B (en) | For exporting the method for candidate character strings, computing device and storage medium | |
| CN103853702B (en) | The apparatus and method of the Chinese idiom mistake in correction language material | |
| US20200026488A1 (en) | Coding system and coding method using voice recognition | |
| CN103019407B (en) | Input method application method, automatic question answering processing method, electronic equipment and server | |
| AU2013270485C1 (en) | Input processing method and apparatus | |
| KR102256705B1 (en) | Training acoustic models using modified terms | |
| CN107748784B (en) | Method for realizing structured data search through natural language | |
| CN112417102A (en) | Voice query method, device, server and readable storage medium | |
| WO2009049049A1 (en) | Method and system for adaptive transliteration | |
| CN106325488B (en) | A kind of input method, input unit, server and input system | |
| CN104166462A (en) | Input method and system for characters | |
| EP2698727A2 (en) | Terminal and method for determining type of input method editor | |
| CN105702252A (en) | Voice recognition method and device | |
| CN110738997A (en) | information correction method, device, electronic equipment and storage medium | |
| CN103365573A (en) | Method and device for identifying multi-key input characters | |
| CN101950240A (en) | Pinyin input method for touch screen | |
| CN113268981B (en) | Information processing method and device and electronic equipment | |
| CN112269475A (en) | Character display method and device and electronic equipment | |
| CN102945120A (en) | Children application based man-machine interaction auxiliary system and interaction method | |
| CN102063282B (en) | Chinese speech input system and method | |
| CN111880668A (en) | Input display method and device and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130313 |