CN102722483B

CN102722483B - For determining method, device and equipment that the candidate item of input method sorts

Info

Publication number: CN102722483B
Application number: CN201110076480.6A
Authority: CN
Inventors: 孟可丰
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2011-03-29
Filing date: 2011-03-29
Publication date: 2017-07-25
Anticipated expiration: 2031-03-29
Also published as: CN102722483A

Abstract

The present invention provides a kind of method, device and the equipment of the candidate item sequence for being used to determine input method.Wherein, the collator first obtains the list entries from user；Then further according to the list entries, inquired about in dictionary, to obtain the multiple candidate items matched with the list entries, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item, to be ranked up processing to the multiple candidate item, to obtain the ranking results of the sequence processing；Finally further according to the ranking results, generation shows information to be supplied to the user.Advantages of the present invention includes：Enable to the sequence of the acquired candidate item matched with list entries more reasonable, be easy to user quickly to select the candidate item needed for it, improve the input speed of user.

Description

For determining method, device and equipment that the candidate item of input method sorts

Technical field

The present invention relates to computer realm, more particularly to a kind of method for the candidate item sequence for determining input method, device and Equipment.

Background technology

With the popularization of computer technology, computer equipment has been entered into the work, study and life of people.For example, People write document, editor's manuscript using computer equipment；In another example, people by the connection of computer and network come with a distant place Friend chat, see a film, listen music, play game etc..

In the prior art, user is inputted after list entries by human-computer interaction devices such as keyboards to computer equipment, meter The input method that calculation machine equipment is provided is provided a user corresponding candidate item based on the list entries and selected for user, to realize The function that text is interacted is carried out with user, to allow users to write document and friend's chat etc. by computer equipment.So And, in each candidate item that existing input method is provided, the predetermined candidate item sequence often provided according only to input method is come to user There is provided the candidate item after sequence so that the really necessary candidate item of user often sorts rearward, causes user to be frequently necessary to page turning Required candidate item could be selected, the input speed of user is reduced.

The content of the invention

It is an object of the invention to provide a kind of method, device and the equipment of the candidate item sequence for determining input method.

There is provided a kind of computer implemented candidate item sequence for being used to determine input method according to an aspect of the present invention Method, wherein, this method comprises the following steps：

A obtains the list entries from user；

B is inquired about according to the list entries in dictionary, to obtain the multiple times matched with the list entries Option, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item, to the multiple candidate Item is ranked up processing, to obtain the ranking results of the sequence processing；

C shows information to be supplied to the user according to the ranking results, generation.

According to another aspect of the present invention, a kind of sequence dress for the candidate item sequence for being used to determine input method is additionally provided Put, wherein, the collator includes：

First acquisition device, for obtaining the list entries from user；

Sort processing unit, for according to the list entries, being inquired about in dictionary, with obtain and the input sequence The multiple candidate items matched, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item are arranged, To be ranked up processing to the multiple candidate item, to obtain the ranking results of the sequence processing；

Generating means, for according to the ranking results, generation to show information to be supplied to the user.

According to a further aspect of the invention, a kind of computer equipment is additionally provided, wherein, before the computer equipment includes State collator.

Compared with prior art, the present invention has advantages below：1) by based on the classification belonging to each candidate item, going forward side by side one Step combines other sequence relevant informations and is ranked up, and enables to sequence more reasonable, is easy to needed for user quickly selects it Candidate item, improves the input speed of user；2) can be according to the current input environment of user, the weight of adjustment classification is believed in real time Breath so that the candidate item sequence of classification corresponding with the input environment that user is current is forward, so that the sequence of candidate item more meets The demand of user；3) can be according to new new-added item acquired in the current input environment of the list entries and user that are inputted by user To extend dictionary.

Brief description of the drawings

By reading the detailed description made to non-limiting example made with reference to the following drawings, of the invention is other Feature, objects and advantages will become more apparent upon：

The flow chart for the method that Fig. 1 sorts for the candidate item for determining input method of one aspect of the invention；

The flow chart for the method that Fig. 2 sorts for the candidate item for determining input method of a preferred embodiment of the invention；

The flow for the method that Fig. 3 sorts for the candidate item for determining input method of another preferred embodiment of the present invention Figure；

The collator schematic diagram that Fig. 4 sorts for the candidate item for determining input method of one aspect of the invention；

Fig. 5 illustrates for the collator of the candidate item sequence for being used to determine input method of a preferred embodiment of the invention Figure；

Fig. 6 illustrates for the collator of the candidate item sequence for being used to determine input method of another preferred embodiment of the present invention Figure；

Same or analogous reference represents same or analogous part in accompanying drawing.

Embodiment

The present invention is described in further detail below in conjunction with the accompanying drawings.

Fig. 1 shows the flow chart of the method for the candidate item sequence for being used to determine input method of one aspect of the invention.Its In, the method according to the invention is mainly completed by the operating system or processing controller in computer equipment, is concise rise See, the operating system or processing controller are referred to as collator below.Wherein, the computer equipment includes but not limited In：1) user equipment；2) network equipment.The user equipment includes but is not limited to computer, smart mobile phone, PDA etc.；The net Network equipment includes but is not limited to single network server, the server group of multiple webservers composition or based on cloud computing The cloud being made up of a large amount of computers or the webserver of (Cloud Computing), wherein, cloud computing is Distributed Calculation One kind, a super virtual computer being made up of the computer collection of a group loose couplings.

In step S14, the collator obtains the list entries from user.

Wherein, the mode of the list entries of the collator acquisition from user includes but is not limited to：

1) collator obtains the list entries that user is inputted by human-computer interaction device；

For example, the collator obtains the list entries " weibo " of user's input.

Wherein, the human-computer interaction device includes but is not limited to：A) keyboard；B) mouse；C) felt pen；D) remote control；e) Voice-operated device etc..

The input method used during user's input list entries includes but is not limited to：

A) Chinese character coding input method.The Chinese character coding input method includes but is not limited to：I) spelling；Ii) simplicity；Iii) stroke；Iv) five Pen etc..

B) English input method.

2) collator is obtained after the operation that user carries out selection candidate item, not yet determines the input of corresponding candidate Sequence；

For example, the list entries " weibo " inputted for user, the candidate item that collator is provided a user includes " micro- Ripple ", " weber " and " not ", if user's selection candidate item " not ", collator obtains the input for not yet determining corresponding candidate Sequence " bo " is used as acquired list entries.

Then, in step S15, the collator is inquired about according to the list entries in dictionary, to obtain The multiple candidate items matched with the list entries, and the classification correspondence according to belonging to each candidate item in the plurality of candidate item Weight information, to be ranked up processing to the multiple candidate item, with obtain it is described sequence processing ranking results.Wherein, The dictionary includes the dictionary that the collator can be provided.For example, the collator provide be in computer equipment In local dictionary；In another example, the network lexicon being in the network equipment that the collator is provided.For another example the row Cloud dictionary being in cloud that sequence device is provided etc..

Wherein, those skilled in the art should be able to determine according to the list entries in dictionary according to actual conditions and demand It is middle to be inquired about, to obtain the candidate item acquisition modes of the candidate item matched with the list entries.For example, when input method is Chinese all-phonetic input method, and list entries be " shanghai " when, the acquired candidate item bag matched with the list entries Include the same or analogous candidate item of all or part of spelling and the list entries：" Shanghai ", " mountain and sea ", " on " etc..This area It is to be understood by the skilled artisans that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than limitation of the present invention, appoint The acquisition modes for the candidate item that acquisition included in what prior art matches with list entries, should be included in the present invention's In the range of.

Wherein, the collator determines class in acquired multiple candidate items belonging to each candidate item otherwise Including but not limited to：

1) comprising many sub- dictionaries in the dictionary, one classification of each sub- dictionary correspondence, and it is comprising belonging to such Other multiple candidate items, sub- dictionary of the collator according to belonging to candidate item determines the classification of candidate item.For example, sub- word The classification in storehouse is computer classification, then the classification of the candidate item from the sub- dictionary is computer classification etc..Wherein, the candidate Item includes character string and character string combinations, it may for example comprise word, word, portmanteau word, short sentence etc..

2) dictionary includes multiple candidate items, and each candidate item corresponds to one or more classifications, the sequence Device determines the classification of candidate item by the corresponding relation of predetermined candidate item and classification.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any class determined belonging to candidate item is otherwise, should be included in the model of the present invention In enclosing.

Wherein, the weight information includes at least one of following：1) weighted value；2) weighting levels.The collator leads to Cross and obtain the corresponding relation of pre-stored of all categories and weight information come corresponding to the candidate item generic acquired in determining Weight information.Preferably, the corresponding relation of all categories with weight information can be determined by user's setting.Need explanation It is that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than limitation of the present invention, those skilled in the art should The understanding, the mode of any determination corresponding weight information of all categories, should be included in the scope of the present invention.

Specifically, the collator is inquired about according to the list entries in dictionary, to obtain and the input Multiple candidate items that sequence matches, and the corresponding weight letter of classification according to belonging to each candidate item in the plurality of candidate item Breath, to be ranked up processing to the multiple candidate item, to include in the way of obtaining the ranking results of the sequence processing but not It is limited to：

1) collator is in the dictionary by obtaining after all candidate items matched with list entries, according to being obtained The corresponding weight information of classification belonging to each candidate item in all candidate items taken, to arrange all candidate items Sequence processing.

For example, the collator obtains all candidate items matched with list entries " weibo " in the dictionary Including " weber " and " microwave ".Wherein, the classification belonging to candidate item " weber " is name classification, the corresponding weight of name classification Information is 3, and the classification belonging to candidate item " microwave " is electronics classification, and the corresponding weight information of electronics classification is 2, then institute Collator is stated according to weight information from high to low come the result obtained by being ranked up processing to candidate item " weber " and " microwave " For：

1 weber；

2 microwaves.

2) collator often obtains a candidate item in the dictionary, i.e., be ranked up processing to the candidate item, And determine the candidate item it is all obtained candidate item in sorting position, until acquired candidate item reached predetermined quantity Or untill having traveled through the dictionary and being ranked up processing to last candidate item of acquisition.

For example, the collator is sorted the candidate item " microwave " by being obtained in the dictionary after candidate item " microwave " At first；Then, the collator obtains candidate item " weber " in the dictionary, and based on candidate item " microwave " with waiting The corresponding weight information of classification belonging to option " weber " is respective, by candidate item " weber " sequence before candidate item " microwave "； Then, the collator continues to inquire about in the dictionary, after the 3rd candidate item is obtained, based on the 3rd candidate item with Candidate item " microwave " and candidate item " weber " three each belonging to the corresponding weight information of classification, to be ranked up place to three Reason；Collator repeats the above steps, until acquired candidate item has reached predetermined quantity, or, the dictionary has been traveled through, And processing is ranked up to last candidate item of acquisition untill.

3) collator in the dictionary by obtaining after all candidate items matched with list entries, by acquired All candidate items in select the candidate item of predetermined quantity, and belonging to each candidate item in the candidate item of the predetermined quantity The corresponding weight information of classification, processing is ranked up come the candidate item to the predetermined quantity, to obtain at the sequence The ranking results of reason.Wherein, the mode of the candidate item of the collator selection predetermined quantity includes but is not limited to：1) random choosing Select；2) order that is acquired according to each candidate item is selected；3) the corresponding weight information of classification according to belonging to candidate item To select.

One of preferably, the predetermined quantity is less than or equal in a display page of input method and can shown The maximum quantity for the candidate item shown.For example, the maximum for the candidate item that can be shown in a display page of the input method Quantity is 8, then the collator selection is ranked up processing no more than 8 candidate items.

For example, the collator obtains all candidate item bags matched with list entries " chen " in the dictionary Include " old ", " morning ", " orange ", " scale " and " multiplying ", then, the sequencing that the collator is acquired according to each candidate item 3 candidate items are therefrom selected, for example, candidate item " old ", " morning " and " orange " are selected, and the classification according to belonging to candidate item " old " is Name classification, the name classification respective weights information is the 4th grade, and the classification belonging to candidate item " morning " is natural science classification, The corresponding weight information of natural science classification is the 1st grade, and the classification belonging to candidate item " orange " is food classification, the foodstuff Not corresponding weight information is the 3rd grade, to determine that candidate item " old ", " morning " and the ranking results of " orange " are：

1 is old；

2 oranges；

3 mornings.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any corresponding power of classification according to belonging to each candidate item in the plurality of candidate item Weight information,, all should in the way of obtaining the ranking results of the sequence processing to be ranked up processing to the multiple candidate item Within the scope of the present invention.

Then, in step S 16, the collator shows information to be supplied to according to the ranking results, generation State user.Wherein, collator generation show information mode include it is a variety of, for example：1) exhibition is generated in horizontally-arranged mode Existing information；2) generated in the way of vertical setting of types and show information etc..

For example, the collator is according to the ranking results, the candidate item next life of the preceding predetermined quantity of selected and sorted Show information into horizontally-arranged.Also, the maximum quantity that the predetermined quantity can show for a display page of input method 3, the then collator three candidate items " old ", " orange " and " morning " of 3 before selected and sorted in ranking results, and with as follows Shown horizontally-arranged mode is supplied to the user：

1 old 2 orange 3 morning.

In another example, during vertical setting of types, the maximum quantity that a display page of input method can be shown is 5, the then sequence Device two candidate items " old " of 2 and " orange " before selected and sorted in ranking results, and carried in the way of vertical setting of types as follows Supply the user：

1 is old

2 oranges.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any according to the ranking results, generation shows information to be supplied to the user's Implementation, for example, the display page additional color and background of input method are showed into information to generate；In another example, by candidate item Font adjustment show information etc. to generate, should be included in the scope of the present invention.

One of preferred scheme as the present invention, the method according to the invention is further comprising the steps of：The collator Obtain the renewal request that being used for from the user ask to update candidate item, and update request according to described and described show Information, selects candidate item to be showed in the ranking results, further according to the candidate item to be showed, and shows described in renewal Information.

For example, the collator time of 5 before selected and sorted in the ranking results of 20 candidate items obtained Option shows information and is supplied to the user to generate.The user is being browsed after this shows information, is sent request and is updated and waits The renewal request of option, for example to press, " Page Down " keys send the renewal request that request updates candidate item, the then sequence Device is obtained should be " after the signal of Page Down " keys, according to the candidate showed included in information for having been provided to the user Sorting position in ranking results, in ranking results after 5 candidate item next life before selected and sorted in the candidate item of 15 Cheng Xin's shows information, and is supplied to the user.

One of preferred scheme as the present invention, the method according to the invention is further comprising the steps of：The collator The renewal request that request updates candidate item that is used for from the user is obtained, and is asked and the exhibition when according to described update Existing information judges in the ranking results not comprising when needing the candidate item showed, according in the multiple candidate item each not The corresponding weight information of classification belonging to the candidate item in the ranking results is contained in, to be not included in the sequence to described As a result the candidate item in is ranked up processing, with the ranking results after being updated, and according to the ranking results after the renewal, Show information to be supplied to the user described in updating.

For example, the collator has obtained 30 candidate items, and have determined that the row of 6 candidate items in 30 candidate items Sequence result, and the ranking results generation based on 6 candidate items shows information there is provided to the user.The user is clear Look at that this shows after information, send the renewal request that request updates candidate item, for example, update candidate to click on " ↓ " key to send request Renewal request, then the collator obtain renewal request, and according to having been provided to the showing in information of the user All ordering candidate items are included, judge not include the candidate item for needing to be showed in the ranking results, then the row The corresponding weight information of classification of the sequence device according to belonging to the candidate item being not included in the ranking results, not wrapped to described The candidate item being contained in the ranking results is ranked up processing, the ranking results after being updated with obtaining ranking results, Information is showed with renewal and the user is supplied to.Wherein, the collator is to the time that is not included in the ranking results Option is ranked up in the mode of processing and abovementioned steps S15 collator according to belonging to each candidate item in multiple candidate items The corresponding weight information of classification, it is same or similar come mode that processing is ranked up to the multiple candidate item, no longer go to live in the household of one's in-laws on getting married herein State.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that renewal request of any acquisition from user is to update the implementation of candidate item, Should within the scope of the present invention.

The method according to the invention, by being sorted based on the classification belonging to each candidate item, enables to sequence more to close Reason, is easy to user quickly to select the candidate item needed for it, improves the input speed of user.

Fig. 2 shows the flow of the method for the candidate item sequence for being used to determine input method of a preferred embodiment of the invention Figure.Step S14, step S15 ' and step S16 are included according to the method for the present embodiment.Wherein, step S14 and step S16 exists With reference to being described in detail in embodiment illustrated in fig. 1, and it is incorporated herein, repeats no more by reference.

In step S15 ', the collator is inquired about according to the list entries in the dictionary, to obtain The multiple candidate items matched with the list entries, and the classification correspondence according to belonging to each candidate item in the plurality of candidate item Weight information and other sequence relevant informations, to be ranked up processing to the multiple candidate item, to obtain the sequence The ranking results of processing.

Wherein, other described sequence relevant informations include at least one of following：

1) each corresponding word frequency information of candidate item in the multiple candidate item.Wherein, the collator determines candidate The mode of the corresponding word frequency information of item includes but is not limited at least one of following：A) frequency of the candidate item is selected according to the user Rate is determined；B) the multiple users counted according to input method select the frequency of the candidate item to determine；C) according to the user The frequency of the candidate item set is determined；D) according to it is foregoing a), b) and c) at least two determine each candidate item correspondence Word frequency；E) the corresponding word frequency information of default candidate item is obtained.

2) each matching degree between candidate item and the list entries in the multiple candidate item.Specifically, the sequence The input method that is used is determined between each candidate item and the list entries when device carries out input operation according to the user Matching degree.For example, when the user uses all-phonetic input method list entries " baidu ", then the collator is according to acquired The phonetic matching degree to determine " crossing in vain " and list entries identical with the list entries of candidate item " crossing in vain " be 1, according to institute The phonetic of the candidate item " white " of acquisition is identical with " bai " in the list entries to determine that " white " and the matching degree of the list entries be 0.5.In another example, when the user uses Zheng's code inputting method list entries " ld ", then the collator determines the candidate obtained The matching degree of item " same " and the list entries is higher than candidate item " door " and the matching degree of the list entries.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any implementation for determining the corresponding word frequency information of candidate item and any determination The implementation of matching degree between candidate item and list entries, should be included in the scope of the present invention.

Wherein, classification corresponding weight information of the collator according to belonging to each candidate item in multiple candidate items with And other sequence relevant informations, include but is not limited to come processing mode that processing is ranked up to the multiple candidate item：

1) the first corresponding weight information of classification according to belonging to each candidate item of the collator or other described sequences One in relevant information is ranked up processing to the multiple candidate item, to obtain first ranking results, then further according to The other information for being not yet used for sequence processing in the weight information and/or other described sequence relevant informations is come to first minor sort As a result it is adjusted.

For example, the candidate item that the collator is obtained includes candidate item A, candidate item B and candidate item C, and it is acquired The corresponding weight information of classification belonging to candidate item A is the 3rd grade, and word frequency information is 5, and the classification belonging to candidate item B is corresponding Weight information is the 3rd grade, and word frequency information is 6, and the corresponding weight information of the classification belonging to candidate item C is the 1st grade, word frequency letter Cease for 2；Then the collator is first ranked up based on the corresponding weight information of classification belonging to candidate item to three candidate items Handle, and obtain first ranking results and be：Candidate item A and candidate item B sorts before candidate item C；Then, the collator Determine that the candidate item A and candidate item B with equal weight information is ordered as based on word frequency information again：Candidate item B sequences are being waited Before option A；Then the collator is ranked up the sequence knot obtained after processing to candidate item A, candidate item B and candidate item C It is really as follows：

1st, candidate item B；

2nd, candidate item A；

3rd, candidate item C.

2) classification corresponding weight information of the collator directly according to belonging to each candidate item in multiple candidate items And other sequence relevant informations, to be ranked up to the multiple candidate item.

For example, the collator is first based on the corresponding weight information of classification belonging to each candidate item and other rows Sequence relevant information, obtains the evaluation of estimate of the candidate item, then the evaluation of estimate based on each candidate item to sort.For example, the row Candidate item acquired in sequence device includes candidate item D and candidate item E, and the corresponding power of classification belonging to acquired candidate item D Weight information is 2, and word frequency information is 3, and the matching degree with list entries is 0.5, the corresponding weight information of classification belonging to candidate item E For 3, word frequency information is 1, and the matching degree with list entries is 0.8, then the collator by weight information, word frequency information and Product with degree is used as the evaluation of estimate of candidate item, and the evaluation of estimate for obtaining candidate item D is 3, and candidate item E evaluation of estimate is 2.4；Then The evaluating apparatus is ordered as based on evaluation of estimate to candidate item D with candidate item E again：

1st, candidate item D；

2nd, candidate item E.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any corresponding weight of classification according to belonging to each candidate item in multiple candidate items Information and other sequence relevant informations, to be ranked up processing to the multiple candidate item, to obtain the sequence processing The implementation of ranking results, for example, after weight information and other sequence relevant informations are normalized, by items letter Cease the value obtained by normalization and, quadratic sum, product or logarithm and be used as evaluation of estimate etc., should be included in the scope of the present invention It is interior.

One of preferred scheme as the present invention, the embodiment shown in step S15 and Fig. 2 in the embodiment shown in Fig. 1 In step S15 ' can be replaced by step S15 " (being unillustrated).

In step S15 ", the collator according to the list entries, the dictionary and with the user institute Inquired about in the program dictionary that the related application program of the input operation of execution is provided, to obtain and the list entries phase The multiple candidate items matched somebody with somebody, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item, to institute State multiple candidate items and be ranked up processing, to obtain the ranking results of the sequence processing.Wherein, the collator determine by The classification belonging to candidate item and the mode of the corresponding weight information of the category obtained in program dictionary includes but is not limited to：Directly Connect the determination candidate item and belong to a predetermined classification, for example, belong to program word class, and determine the corresponding weight of the predetermined classification Information is most high.

For example, when the user inputs list entries " wangfei ", the sequence dress in application program " 1,000 hark " Put based on the acquired list entries, inquired about in the dictionary, obtain candidate item " imperial concubine ", and " 1,000 hark " Inquired about in the program dictionary of offer, obtain candidate item " Wang Fei ", then the collator is based on belonging to candidate item " Wang Fei " The corresponding weight information of classification be peak 10, and the corresponding weight information of classification belonging to candidate item " imperial concubine " is 5, it is determined that Ranking results are as follows：

1st, Wang Fei；

2nd, imperial concubine.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that it is any determine as the classification belonging to the candidate item obtained in program dictionary and such The implementation of not corresponding weight information, should be included in the scope of the present invention.

One of preferred scheme as the present invention, the method according to the invention is further comprising the steps of：The collator The selected candidate item from described program dictionary of the user is added in the dictionary.

For example, when the user comes from the middle input list entries " wangfei " of application program " 1,000 hark ", and selection Candidate item " Wang Fei " in the program dictionary that application program " 1,000 hark " is provided, and when collator determines the candidate item " king It is luxuriant and rich with fragrance " be not included in the dictionary, then the collator adds the candidate item " Wang Fei " in the dictionary.

The method according to the invention, by being ranked up based on the classification belonging to candidate item and other sequence relevant informations, Sequence can be made more to meet the demand of user, user is more quickly selected required candidate item, the input speed of user is improved.

Fig. 3 shows the stream of the method for the candidate item sequence for being used to determine input method of another preferred embodiment of the present invention Cheng Tu.Step S11, step S12, step S13, step S14, step S15 " ' and step are included according to the method for the present embodiment S16。

In step s 11, the collator obtains the current application related to the input operation performed by the user Environmental information.

Wherein, the current application environmental information includes at least one of following：

1) application information related to the input operation.Wherein, the application program related to the input operation Information includes but is not limited to start identification information of application program of input method etc..For example, when the user is in application program " evil spirit List entries is inputted in the beast world ", then the collator obtains the application information bag related to user's input operation Include the identification information for the application program " World of Warcraft " for starting input method.Preferably, collator can be by such as obtaining loading The identification information of the application program of input method, or, the corresponding application program of current active window is obtained in system information The modes such as identification information determine the application information related to the input operation.

2) the program scene information related to the input operation.Wherein, the program scene related to the input operation Information includes but is not limited to scene identity information corresponding to cursor position etc..For example, when the user is in application program The addressee column of " outlook " inputs list entries, then the collator obtains the program related to the input operation Scape information includes the identification information on " the addressee column " corresponding to cursor position.Preferably, collator can be by such as obtaining Input method is currently located the modes such as the attribute of input frame to determine the program scene information related to the input operation.

3) text message included in the application scenarios related to the input operation.Wherein, grasped with the input The text message for making to include in related application scenarios include but is not limited to cursor where the text that includes of the display page Information etc..For example, when inputting list entries in Chat page of the user at one QQ groups, then acquired in the collator The application scenarios related to the input operation in the text message that includes include the chat page of the QQ groups where cursor Group's name information of the QQ groups that face is included, the user of all participations shown in the Chat page of the QQ groups chat chat The text messages such as its content information.In another example, when input input sequence in the display page of the user in application program " word " During row, the text message included in the application scenarios related to the input operation acquired in the collator includes The text messages such as heading message and content information that the word Page Names information where cursor, the word pages are included.It is excellent Selection of land, collator can by such as obtaining the text message that input method is currently located in input frame, or, currently should obtain The program scene letter related to the input operation is determined with the mode such as program and the text description information of program scene information Breath.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any current application environment letter related to input operation performed by the user Breath, should be included in the scope of the present invention.

Then, in step s 12, the collator determines the associated class related to the current application environmental information Not.

Specifically, collator one or more directly in the current application environmental information is predetermined current Inquired about in the corresponding relation of application environment information and classification, to obtain and one or more in current application environmental information Related related category；Or, collator first carries out respective handling to one or more in current application environmental information, then Result according to obtained by processing is inquired about in the corresponding relation of predetermined result and classification, with acquisition and current application One or more of related related category in environmental information.

For example, the mark that the current application environmental information that the collator is obtained includes application program " 1,000 hark " is believed Breath, then the collator is searched in the corresponding relation of predetermined application program identification information and related category, it is determined that The related category related to the identification information of application program " 1,000 hark " includes name classification and song title classification.In another example, The current application environmental information that the collator is obtained includes the identification information on " addressee column ", then the collator is pre- Searched in fixed corresponding relation, it is determined that the related category related to the identification information on " addressee column " includes email address class Other and name classification.For another example the current application environmental information that the collator is obtained includes group's title " piano love of QQ groups Good person group ", then after the collator is to " pianophile group " progress cutting word and duplicate removal processing, by the keyword " steel of acquisition Qin " and " fan " are searched in the corresponding relation of predetermined keyword and related category, it is determined that with this group of name informations Related related category includes piano classification.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any realization for determining the related category related to the current application environmental information Mode, for example, when the corresponding relation is set up by the model of pre-training, collator determines text according to the model of pre-training The classification of this information, enters to be about to the classification of text message as related category etc., should be included in the scope of the present invention.

Then, in step s 13, the collator performs according to the current application environmental information and adjusts the correlation The adjustment operation of the corresponding weight information of classification.Wherein, the collator adjusts the corresponding weight information of the related category Mode include but is not limited to：Weight information is heightened into predetermined value, predetermined grade is heightened, adjusts to peak, adjusts to highest Grade etc..

For example, the mark that the current application environmental information that the collator is obtained includes application program " 1,000 hark " is believed Breath, it is determined that the related category related to the identification information of the application program " 1,000 hark " includes name classification and song title class Not, then the collator adjusts name classification and each self-corresponding weight information of song title classification to peak.

Step S14 is described in detail with reference to the embodiment shown in FIG. 1, and is incorporated herein by reference, is no longer gone to live in the household of one's in-laws on getting married State.It should be noted that step S14 and abovementioned steps S11 to step S13 has no sequencing.

In step S15 " ', the collator is inquired about according to the list entries in the dictionary, to obtain The multiple candidate items matched with the list entries are taken, and according to each in the plurality of candidate item after the execution adjustment operation The corresponding weight information of classification belonging to candidate item, to be ranked up processing to the multiple candidate item, to obtain the sequence The ranking results of processing.The collator is according to belonging to performing each candidate item in the plurality of candidate item after the adjustment operation The corresponding weight information of classification, to be ranked up the mode of processing and the implementation shown in earlier figures 1 to the multiple candidate item Classification corresponding weight information of the collator according to belonging to each candidate item in multiple candidate items in step S15 in example, comes The mode for being ranked up processing to the multiple candidate item is same or similar, will not be repeated here.

One of preferred scheme as the present invention, the method according to the invention is further comprising the steps of：The collator Determine the second new new-added item according to the text message, and the described second new new-added item is added in the dictionary, and by the phase Classification is closed as the classification of the described second new new-added item.

For example, the heading message included in the word pages that the collator is obtained includes " computer anti-virus skill Content information included in art ", the word pages includes：" source code virus ", the collator is according to the acquired mark Inscribe the keyword " virus " in information and determine that the related category related to the current application environmental information includes computer virus Classification, and semantic analysis is carried out to content information " source code virus ", keyword " source code type " and " virus " is obtained, then the row Sequence device is added the keyword being not included in the dictionary " source code type " as the second new new-added item in the dictionary, and will meter Calculation machine virus classification as the new new-added item " source code type " classification.

In another example, the QQ groups name information that the collator is obtained includes " pianophile group ", chat content information Including " I likes Liang shanbo and Zhu yingtai ", the keyword " piano " that the collator is based in " pianophile group " is determined and described current The related related category of application environment information includes piano classification, and by the chat content information, " I likes based on sentence pattern analysis Liang shanbo and Zhu yingtai " are divided into subject " I ", predicate " liking " and object " Liang shanbo and Zhu yingtai ", then the collator will not yet be contained in the dictionary In object " Liang shanbo and Zhu yingtai " as the second new new-added item, add in the dictionary, and regard piano classification as the second new new-added item " beam Wish " classification.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any implementation for determining the second new new-added item, should be included in the scope of the present invention It is interior.

Then, step S16 is described in detail with reference to the embodiment shown in FIG. 1, and is incorporated herein by reference, no Repeat again.

One of preferred scheme as the present invention, the method according to the invention is further comprising the steps of：The collator Obtain the first new new-added item in the list entries, and will be in the described first new new-added item addition dictionary, and by the correlation Classification as the described first new new-added item classification.

For example, the collator, which obtains the application information related to user's input operation, includes " addressee column " Identification information, then the collator determine the related category of the application information correlation related to user's input operation Including email address classification.When the user inputs list entries in " the addressee column "aaa@bbb.com, then the sequence Device obtains the first new new-added item in the list entriesaaa@bbb.com, and by the first new new-added itemaaa@bbb.comAdd institute In predicate storehouse, and using email address classification as the first new new-added item classification.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any implementation that the first new new-added item is obtained in the list entries, all should be wrapped Containing within the scope of the invention.

It is also understood by those skilled in the art that in embodiment shown in Fig. 3, step S11 to S13 and step S14 is simultaneously Without sequencing, in fact, collator can be first carried out after step S14, then step S11, S12 and S13 are sequentially performed；Or, Collator is after step S11 to step S13 is performed, then sequentially performs step S14 etc.；Or, collator performs step parallel Rapid S14 and step S11 to step S13.

The method according to the invention, can be according to the current input environment of user, and the weight information of adjustment classification, makes in real time The candidate item sequence for obtaining classification corresponding with the input environment that user is current is forward, so that the sequence of candidate item more meets user's Demand, is easy to user quickly to select correct candidate item, improves the input speed of user.

One of preferred scheme as the present invention, the method according to the invention is further comprising the steps of：The collator Frequency is selected according to the history of the candidate item of each classification, to adjust the weight information of each classification.Wherein, each classification is adjusted The mode of weight information include but is not limited to：

1) select the result that frequency is compared to each other to adjust the weight information of each classification based on each history.For example, described The history selection frequency that collator acquisition belongs to computer other candidate item is 800 times, belongs to the candidate item of music categories History selection frequency be 500 times, belong to mechanical other candidate item history selection frequency be 10 times, computer classification correspondence Weight information be 2, the corresponding weight information of music categories is 6, and the corresponding weight information of mechanical classification is 5, the then sequence After more each history selection frequency of device, determine that the history selection frequency of computer other candidate item is higher than the time of music categories The history selection frequency of option, the history selection frequency of the candidate item of music categories is higher than the history of mechanical other candidate item again Frequency is selected, thus, the corresponding weight information of computer classification is adjusted to corresponding higher than music categories by the collator Weight information 7.

2) comparative result of frequency and predetermined threshold is selected to adjust the weight information of each classification based on history.For example, When the collator judges that the history selection frequency of the candidate item of a classification is higher than predetermined threshold, then the collator will The corresponding weight information of the category is heightened；In another example, when the collator judges the history selection of the candidate item of a classification Frequency is less than predetermined threshold, then the collator turns down the corresponding weight information of the category.Wherein, the width heightened or turned down Spend for predetermined amplitude.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that the history selection frequency of any candidate item according to each classification, to adjust each class The implementation of other weight information, should be included in the scope of the present invention.

Fig. 4 shows the signal of the collator of the candidate item sequence for being used to determine input method of one aspect of the invention Figure.Wherein, the first acquisition device 11, sequence processing unit 12 and generating means 13 are included according to the collator 1 of the present embodiment.

First acquisition device 11 obtains the list entries from user.

Wherein, the mode that first acquisition device 11 obtains the list entries from user includes but is not limited to：

1) first acquisition device 11 obtains the list entries that user is inputted by human-computer interaction device；

For example, first acquisition device 11 obtains the list entries " weibo " of user's input.

B) English input method.

2) first acquisition device 11 is obtained after the operation that user carries out selection candidate item, not yet determines corresponding candidate List entries；

For example, the list entries " weibo " inputted for user, the candidate item that collator is provided a user includes " micro- Ripple ", " weber " and " not ", if user's selection candidate item " not ", the first acquisition device 11, which is obtained, not yet determines corresponding candidate List entries " bo " be used as acquired list entries.

Then, the sequence processing unit 12 is inquired about, with acquisition and institute according to the list entries in dictionary 2 State multiple candidate items that list entries matches, and the corresponding power of classification according to belonging to each candidate item in the plurality of candidate item Weight information, to be ranked up processing to the multiple candidate item, to obtain the ranking results of the sequence processing.Wherein, it is described Dictionary 2 includes the dictionary that the sequence processing unit 12 can be provided.For example, being in for sequence processing unit 12 offer is counted Calculate the local dictionary in machine equipment；In another example, the network lexicon being in the network equipment that the sequence processing unit 12 is provided. For another example cloud dictionary being in cloud that the sequence processing unit 12 is provided etc..

Wherein, the sequence processing unit 12 determines the classification in acquired multiple candidate items belonging to each candidate item Mode include but is not limited to：

1) comprising many sub- dictionaries in the dictionary 2, one classification of each sub- dictionary correspondence, and it is comprising belonging to such Other multiple candidate items, sub- dictionary of the sequence processing unit 12 according to belonging to candidate item determines the classification of candidate item.Example Such as, the classification of sub- dictionary is computer classification, then the classification of the candidate item from the sub- dictionary is computer classification etc..Wherein, The candidate item includes character string and character string combinations, it may for example comprise word, word, portmanteau word, short sentence etc..

2) dictionary 2 includes multiple candidate items, and each candidate item corresponds to one or more classifications, the sequence Processing unit 12 determines the classification of candidate item by the corresponding relation of predetermined candidate item and classification.

Wherein, the weight information includes at least one of following：1) weighted value；2) weighting levels.The sequence processing dress Candidate item generic institute acquired in putting 12 and being determined by obtaining the corresponding relation of pre-stored of all categories and weight informations Corresponding weight information.Preferably, the corresponding relation of all categories with weight information can be determined by user's setting.Need Illustrate, the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than limitation of the present invention, art technology Personnel are it should be understood that the mode of any determination corresponding weight information of all categories, should be included in the scope of the present invention.

Specifically, it is described sequence processing unit 12 according to the list entries, inquired about in dictionary 2, with obtain and Multiple candidate items that the list entries matches, and the classification according to belonging to each candidate item in the plurality of candidate item is corresponding Weight information, to be ranked up processing to the multiple candidate item, the bag in the way of obtaining the ranking results of the sequence processing Include but be not limited to：

1) the sequence processing unit 12 is in the dictionary 2 by obtaining after all candidate items matched with list entries, The corresponding weight information of classification belonging to each candidate item in acquired all candidate items, to all candidates Item is ranked up processing.

For example, the sequence processing unit 12 is obtained and owning that list entries " weibo " matches in the dictionary 2 Candidate item includes " weber " and " microwave ".Wherein, the classification belonging to candidate item " weber " is name classification, name classification correspondence Weight information be 3, classification belonging to candidate item " microwave " is electronics classification, and the corresponding weight information of electronics classification is 2, then the sequence processing unit 12 place is ranked up to candidate item " weber " and " microwave " according to weight information from high to low Result obtained by reason is：

1 weber；

2 microwaves.

2) the sequence processing unit 12 often obtains a candidate item in the dictionary 2, i.e., the candidate item is arranged Sequence processing, and determine the candidate item it is all obtained candidate item in sorting position, until acquired candidate item reached Predetermined quantity or untill having traveled through the dictionary 2 and being ranked up processing to last candidate item of acquisition.

For example, the sequence processing unit 12 in the dictionary 2 by obtaining after candidate item " microwave ", the candidate item is " micro- Ripple " sorts at first；Then, the sequence processing unit 12 obtains candidate item " weber " in the dictionary 2, based on candidate Item " microwave " weight information corresponding with the respective affiliated classification of candidate item " weber ", by candidate item " weber " sequence in candidate item Before " microwave "；Then, the sequence processing unit 12 continues to inquire about in the dictionary 2, after the 3rd candidate item is obtained, Based on the 3rd candidate item and candidate item " microwave " and candidate item " weber " three each belonging to the corresponding weight information of classification, To be ranked up processing to three；Sequence processing unit 12 repeat the above steps, until acquired candidate item reached it is predetermined Quantity, or, travel through the dictionary, and untill being ranked up processing to last candidate item of acquisition.

3) the sequence processing unit 12 is in the dictionary 2 by obtaining after all candidate items matched with list entries, The candidate item of predetermined quantity is selected in acquired all candidate items, and according to each time in the candidate item of the predetermined quantity The corresponding weight information of classification belonging to option, is ranked up processing, to obtain come the candidate item to the predetermined quantity State the ranking results of sequence processing.Wherein, it is described sequence processing unit 12 selection predetermined quantity candidate item mode include but It is not limited to：1) randomly choose；2) order that is acquired according to each candidate item is selected；3) classification according to belonging to candidate item Corresponding weight information is selected.

One of preferably, the predetermined quantity is less than or equal in a display page of input method and can shown The maximum quantity for the candidate item shown.For example, the maximum for the candidate item that can be shown in a display page of the input method Quantity is 8, then sequence processing unit 12 selection is ranked up processing no more than 8 candidate items.

For example, the sequence processing unit 12 is obtained and owning that list entries " chen " matches in the dictionary 2 Candidate item includes " old ", " morning ", " orange ", " scale " and " multiplying ", then, and the sequence processing unit 12 is according to each candidate item quilt The sequencing of acquisition therefrom selects 3 candidate items, for example, selection candidate item " old ", " morning " and " orange ", and according to candidate item Classification belonging to " old " is name classification, and the name classification respective weights information is the 4th grade, the classification belonging to candidate item " morning " For natural science classification, the corresponding weight information of natural science classification is the 1st grade, and the classification belonging to candidate item " orange " is food Not, the corresponding weight information of food classification is the 3rd grade to category, to determine candidate item " old ", " morning " and the sequence knot of " orange " It is really：

1 is old；

2 oranges；

3 mornings.

Then, the generating means 13 are according to the ranking results, and generation shows information to be supplied to the user.Its In, the generating means 13 generate show information mode include it is a variety of, for example：1) generated in horizontally-arranged mode and show information； 2) generated in the way of vertical setting of types and show information etc..

For example, the generating means 13 are according to the ranking results, the candidate item of the preceding predetermined quantity of selected and sorted is come Generation is horizontally-arranged to show information.Also, the maximum number that the predetermined quantity can show for a display page of input method Amount 3, then the generating means 13 three candidate items " old ", " orange " and " morning " of 3 before selected and sorted in ranking results, and with Horizontally-arranged mode as follows is supplied to the user：

1 old 2 orange 3 morning.

In another example, during vertical setting of types, the maximum quantity that a display page of input method can be shown is 5, the then generation Device 13 two candidate items " old " of 2 and " orange " before selected and sorted in ranking results, and in the way of vertical setting of types as follows It is supplied to the user：

1 is old

2 oranges.

One of preferred scheme as the present invention, the collator 1 also includes the 4th acquisition device (being unillustrated), choosing Select device (being unillustrated) and the first updating device (being unillustrated).

Specifically, the 4th acquisition device obtains the renewal request that request updates candidate item that is used for from the user Afterwards, the selection device according to it is described update request and it is described show information, select to be showed in the ranking results Candidate item, then, first updating device according to the candidate item to be showed, update described in show information.

For example, the generating means 13 are 5 before selected and sorted in the ranking results of 20 candidate items obtained Candidate item shows information and is supplied to the user to generate.The user is being browsed after this shows information, is sent request and is updated The renewal request of candidate item, such as to press " Page Down " keys, which send request and update the renewal of candidate item, asks, then and described the Four acquisition device are obtained should " after the signal of Page Down " keys, the selection device be according to having been provided to showing for the user Sorting position of the candidate item in ranking results included in information, the row of selection in latter 15 candidate item in ranking results 5 candidate items before sequence, then, first updating device generate the new information that shows according to preceding 5 candidate items, and provide To the user.

One of preferred scheme as the present invention, the collator 1 also includes the 5th acquisition device (being unillustrated), son Processing unit (being unillustrated) and the second updating device (being unillustrated).

Specifically, the 5th acquisition device obtains the renewal for being used for request renewal candidate item from the user and asked Ask；Then, the sub-processing unit is judged in the ranking results not when according to the renewal request and the information that shows During comprising needing the candidate item showed, according to the candidate being each not included in the multiple candidate item in the ranking results The corresponding weight information of classification belonging to, to be ranked up place to the candidate item being not included in the ranking results Reason, with the ranking results after being updated；Then, the second updating device updates described according to the ranking results after the renewal Show information to be supplied to the user.

For example, the sequence processing unit 12 has obtained 30 candidate items, and have determined that 6 candidates in 30 candidate items The ranking results of item, the generating means 13 ranking results generation based on 6 candidate items shows information, and there is provided to described User.The user is being browsed after this shows information, sends the renewal request that request updates candidate item, such as to click on " ↓ " key To send the renewal request that request updates candidate item, then the 5th acquisition device is obtained after renewal request, the subprocessing Device includes all ordering candidate items according to showing for the user is had been provided in information, judge the ranking results In not comprising the candidate item that is showed is needed, then the sub-processing unit is according to the candidate item being not included in the ranking results The corresponding weight information of affiliated classification, to be ranked up processing to the candidate item being not included in the ranking results, Ranking results after being updated with obtaining ranking results, then, second updating device is according to the sequence knot after renewal Fruit shows information and is supplied to the user to update.Wherein, the sub-processing unit is to being not included in the ranking results Candidate item be ranked up the mode of processing with it is foregoing sequence processing unit 12 according to belonging to each candidate item in multiple candidate items The corresponding weight information of classification, it is same or similar come mode that processing is ranked up to the multiple candidate item, herein no longer Repeat.

According to the collator of the present invention, by being sorted based on the classification belonging to each candidate item, sequence is enabled to more To be reasonable, it is easy to user quickly to select the candidate item needed for it, improves the input speed of user.

Fig. 5 shows the collator of the candidate item sequence for being used to determine input method of a preferred embodiment of the invention Schematic diagram.Wherein, include the first acquisition device 11, sequence processing unit 12 and generation according to the collator 1 of the present embodiment to fill Put 13；The sequence processing unit 12 includes the first sub- collator 121.

First acquisition device 11 has been described in detail with reference to the embodiment shown in FIG. 4 with generating means 13, and with reference Mode is incorporated herein, and is repeated no more.

The first sub- collator 121 is inquired about according to the list entries in the dictionary 2, with obtain with Multiple candidate items that the list entries matches, and the classification according to belonging to each candidate item in the plurality of candidate item is corresponding Weight information and other sequence relevant informations, to be ranked up processing to the multiple candidate item, to obtain at the sequence The ranking results of reason.

1) each corresponding word frequency information of candidate item in the multiple candidate item.Wherein, the first word collator 121 Determine the corresponding word frequency information of candidate item mode include but is not limited to it is at least one of following：A) time is selected according to the user The frequency of option is determined；B) the multiple users counted according to input method select the frequency of the candidate item to determine；C) basis The frequency of the candidate item that the user is set is determined；D) according to it is foregoing a), b) and c) at least two determine each time The corresponding word frequency of option；E) the corresponding word frequency information of default candidate item is obtained.

2) each matching degree between candidate item and the list entries in the multiple candidate item.Specifically, described first The input method that is used determines each candidate item and the input when sub- collator 121 carries out input operation according to the user Matching degree between sequence.For example, when the user uses all-phonetic input method list entries " baidu ", then first son sorts Device 121 determines " crossing in vain " and list entries according to the phonetic of acquired candidate item " crossing in vain " is identical with the list entries Matching degree is 1, determines that " white " and this are defeated according to the phonetic of acquired candidate item " white " is identical with " bai " in the list entries The matching degree for entering sequence is 0.5.In another example, when the user uses Zheng's code inputting method list entries " 1d ", then described first is sub Collator 121 determines the matching degree of candidate item " same " and the list entries obtained higher than candidate item " door " and the list entries Matching degree.

Wherein, classification of the described first sub- collator 121 according to belonging to each candidate item in multiple candidate items is corresponding Weight information and other sequence relevant informations, include but not to be ranked up the processing mode of processing to the multiple candidate item It is limited to：

1) the classification corresponding weight information or described of the described first sub- collator 121 first according to belonging to each candidate item One in other sequence relevant informations is ranked up processing to the multiple candidate item, to obtain first ranking results, connects The other information further according to the processing that is not yet used to sorting in the weight information and/or other described sequence relevant informations is come pair First ranking results are adjusted.

For example, the candidate item that the first sub- collator 121 is obtained includes candidate item A, candidate item B and candidate item C, and The corresponding weight information of classification belonging to acquired candidate item A is the 3rd grade, and word frequency information is 5, the class belonging to candidate item B Not corresponding weight information is the 3rd grade, and word frequency information is 6, and the corresponding weight information of the classification belonging to candidate item C is 1st etc. Level, word frequency information is 2；Then the described first sub- collator 121 is first based on the corresponding weight information pair of classification belonging to candidate item Three candidate items are ranked up processing, and obtain first ranking results and be：Candidate item A and candidate item B sort candidate item C it Before；Then, the collator determines that candidate item A with equal weight information and candidate item B is arranged based on word frequency information again Sequence is：Candidate item B sorts before candidate item A；Then the described first sub- collator 121 is to candidate item A, candidate item B and candidate It is as follows that item C is ranked up the ranking results obtained after processing：

1st, candidate item B；

2nd, candidate item A；

3rd, candidate item C.

2) classification of the described first sub- collator 121 directly according to belonging to each candidate item in multiple candidate items is corresponding Weight information and other sequence relevant informations, to be ranked up to the multiple candidate item.

Specifically, the first sub- collator 121 is first based on the corresponding weight letter of classification belonging to each candidate item Breath and other sequence relevant informations, obtain the evaluation of estimate of the candidate item, then the evaluation of estimate based on each candidate item to sort. For example, the candidate item acquired in the first sub- collator 121 includes candidate item D and candidate item E, and acquired candidate item The corresponding weight information of classification belonging to D is 2, and word frequency information is 3, and the matching degree with list entries is 0.5, belonging to candidate item E The corresponding weight information of classification be 3, word frequency information is 1, is 0.8 with the matching degree of list entries, then the first son sequence The product of weight information, word frequency information and matching degree is used as the evaluation of estimate of candidate item by device 121, obtains candidate item D evaluation It is worth for 3, candidate item E evaluation of estimate is 2.4；Then the evaluating apparatus is sorted based on evaluation of estimate to candidate item D and candidate item E again For：

1st, candidate item D；

2nd, candidate item E.

As one of preferred scheme of the present invention, sequence processing unit 12 in collator 1 and Fig. 5 institutes shown in Fig. 4 The first sub- collator 121 in the collator 1 shown can include the 3rd sub- collator (being unillustrated).

The 3rd sub- collator according to the list entries, the dictionary 2 and with performed by the user Inquired about in the program dictionary that the related application program of input operation is provided, to obtain and that the list entries matches is many Individual candidate item, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item, to the multiple Candidate item is ranked up processing, to obtain the ranking results of the sequence processing.Wherein, the described 3rd sub- collator determine by The classification belonging to candidate item and the mode of the corresponding weight information of the category obtained in program dictionary includes but is not limited to：Directly Connect the determination candidate item and belong to a predetermined classification, for example, belong to program word class, and determine the corresponding weight of the predetermined classification Information is most high.

For example, when the user inputs list entries " wangfei ", the 3rd son in application program " 1,000 hark " Collator is inquired about based on the acquired list entries in the dictionary 2, obtains candidate item " imperial concubine ", and " thousand Thousand hark " inquired about in the program dictionary that provides, candidate item " Wang Fei " is obtained, then the 3rd sub- collator is based on candidate The corresponding weight information of classification belonging to " Wang Fei " is peak 10, and the corresponding weight of classification belonging to candidate item " imperial concubine " Information is 5, determines that ranking results are as follows：

1st, Wang Fei；

2nd, imperial concubine.

One of preferred scheme as the present invention, the collator 1 also includes the 3rd adding set (being unillustrated).Institute The 3rd adding set is stated to add the user selected candidate item from described program dictionary in the dictionary.

For example, when the user comes from the middle input list entries " wangfei " of application program " 1,000 hark ", and selection Candidate item " Wang Fei " in the program dictionary that application program " 1,000 hark " is provided, and when the 3rd adding set determines the candidate item " Wang Fei " is not included in the dictionary 2, then the 3rd adding set adds the candidate item " Wang Fei " in the dictionary 2.

According to the collator of the present invention, it is ranked up based on the classification belonging to candidate item and other sequence relevant informations, Sequence can be made more to meet the demand of user, user is more quickly selected required candidate, improve the input speed of user.

Fig. 6 shows the collator of the candidate item sequence for being used to determine input method of another preferred embodiment of the present invention Schematic diagram.Wherein, the first acquisition device 11, sequence processing unit 12, generation dress are included according to the collator 1 of the present embodiment Put the 13, second acquisition device 14, the first determining device 15 and the first adjusting apparatus 16；The sequence processing unit 12 is further wrapped Include the second sub- collator 122.

Second acquisition device 14 obtains the current application environment letter related to the input operation performed by the user Breath.Wherein, the current application environmental information includes at least one of following：

1) application information related to the input operation.Wherein, the application program related to the input operation Information includes but is not limited to start identification information of application program of input method etc..For example, when the user is in application program " evil spirit List entries is inputted in the beast world ", then second acquisition device 14 obtains the application program related to user's input operation Information includes starting the identification information of the application program " World of Warcraft " of input method.Preferably, the second acquisition device 14 can pass through The identification information for the application program for being loaded into input method is such as obtained, or, current active window correspondence is obtained in system information The mode such as the identification information of application program determine the application information related to the input operation.

2) the program scene information related to the input operation.Wherein, the program scene related to the input operation Information includes but is not limited to scene identity information corresponding to cursor position etc..For example, when the user is in application program The addressee column input list entries of " outlook ", then the acquisition of the second acquisition device 14 is related to the input operation Program scene information includes the identification information on " the addressee column " corresponding to cursor position.Preferably, the second acquisition device 14 can The mode such as attribute of input frame, which is currently located, by obtaining input method determines the program related to the input operation Scape information.

3) text message included in the application scenarios related to the input operation.Wherein, grasped with the input The text message for making to include in related application scenarios include but is not limited to cursor where the text that includes of the display page Information etc..For example, when inputting list entries in Chat page of the user at one QQ groups, then second acquisition device 14 The text message included in the acquired application scenarios related to the input operation includes the QQ groups' where cursor Group's name information of the QQ groups that Chat page is included, the use of all participations shown in the Chat page of the QQ groups chat The text messages such as the chat content information at family.In another example, when input in the display page of the user in application program " word " During list entries, included in the application scenarios related to the input operation acquired in second acquisition device 14 Heading message and content information that text message is included including the word Page Names information where cursor, the word pages etc. Text message.Preferably, the second acquisition device 14 can be by such as obtaining the text envelope that input method is currently located in input frame Breath, or, the mode such as current application program and the text description information of program scene information is obtained to determine to grasp with the input Make related program scene information.

Then, first determining device 15 determines the related category related to the current application environmental information.

Specifically, one or more directly in the current application environmental information of the first determining device 15 is predetermined Current application environmental information and classification corresponding relation in inquired about, with obtain with current application environmental information in one Or multinomial related related category；Or, the first determining device 15 is first entered to one or more in current application environmental information Row respective handling, is inquired about further according to the result obtained by processing in the corresponding relation of predetermined result and classification, with Obtain and one or more of related related category in current application environmental information.

For example, the current application environmental information that second acquisition device 14 is obtained includes application program " 1,000 hark " Identification information, then first determining device 15 enter in the corresponding relation of predetermined application program identification information and related category Row is searched, it is determined that the related category related to the identification information of application program " 1,000 hark " includes name classification and song title class Not.In another example, the current application environmental information that second acquisition device 14 is obtained includes the identification information on " addressee column ", then First determining device 15 is searched in predetermined corresponding relation, it is determined that related to the identification information on " addressee column " Related category includes email address classification and name classification.For another example the current application ring that second acquisition device 14 is obtained Environment information includes group's title " pianophile group " of QQ groups, then first determining device 15 is carried out to " pianophile group " It is after cutting word and duplicate removal processing, the keyword " piano " of acquisition and " fan " is corresponding with related category in predetermined keyword Searched in relation, it is determined that the related category related to this group of name informations includes piano classification.

It should be noted that the above-mentioned examples are merely illustrative of the technical solutions of the present invention, rather than to the limit of the present invention System, it should be appreciated by those skilled in the art that any realization for determining the related category related to the current application environmental information Mode, for example, when the corresponding relation is set up by the model of pre-training, the first determining device 15 according to the model of pre-training come The classification of text message is determined, enters to be about to the classification of text message as related category etc., should be included in the scope of the present invention It is interior.

Then, first adjusting apparatus 16 performs according to the current application environmental information and adjusts the related category pair The adjustment operation for the weight information answered.Wherein, first adjusting apparatus 16 adjusts the corresponding weight information of the related category Mode include but is not limited to：Weight information is heightened into predetermined value, predetermined grade is heightened, adjusts to peak, adjusts to highest Grade etc..

For example, the current application environmental information that second acquisition device 14 is obtained includes application program " 1,000 hark " Identification information, first determining device determines the related category bag related to the identification information of the application program " 1,000 hark " Name classification and song title classification are included, then first adjusting apparatus 16 is by name classification and each self-corresponding power of song title classification Weight information is adjusted to peak.

Then, first acquisition device 11 is described in detail with reference to the embodiment shown in FIG. 4, and by reference It is incorporated herein, repeats no more.It should be noted that the operation performed by first acquisition device 11 is obtained with foregoing second and filled Put the 14, first determining device 15 and the operation performed by the first adjusting apparatus 16 has no sequencing.

Then, the described second sub- collator 122 is inquired about, to obtain according to the list entries in the dictionary 2 The multiple candidate items matched with the list entries are taken, and according to each in the plurality of candidate item after the execution adjustment operation The corresponding weight information of classification belonging to candidate item, to be ranked up processing to the multiple candidate item, to obtain the sequence The ranking results of processing.The second sub- collator 122 is according to each in the plurality of candidate item after the execution adjustment operation The corresponding weight information of classification belonging to candidate item, to be ranked up the mode and earlier figures 4 of processing to the multiple candidate item Classification corresponding weight information of the shown sequence processing unit 12 according to belonging to each candidate item in multiple candidate items, to institute State multiple candidate items be ranked up processing mode it is same or similar, will not be repeated here.

One of preferred scheme as the present embodiment, the collator 1 also includes the second determining device (being unillustrated) (it is unillustrated) with the second adding set.

Specifically, second determining device is determined according to the text message after the second new new-added item, the second addition dress Put and add the described second new new-added item in the dictionary, and using the related category as the described second new new-added item classification.

For example, the heading message included in the word pages that second acquisition device 122 is obtained includes, " computer is anti- Content information included in virus technology ", the word pages includes：" source code virus ", the basis of the first determining device 15 Keyword " virus " in the acquired heading message determines the related category bag related to the current application environmental information Computer virus classification is included, second determining device carries out semantic analysis to content information " source code virus ", obtain crucial Word " source code type " and " virus ", regard the keyword being not included in the dictionary " source code type " as the second new new-added item, described Two adding sets add the described second new new-added item in the dictionary 2, and regard computer virus classification as the new new-added item " source code The classification of type ".

In another example, the QQ groups name information that second acquisition device 14 is obtained is included in " pianophile group ", chat Holding information includes " I likes Liang shanbo and Zhu yingtai ", and the keyword " piano " that first determining device 15 is based in " pianophile group " is true The fixed related category related to the current application environmental information includes piano classification, and second determining device is based on sentence pattern point The chat content information " I likes Liang shanbo and Zhu yingtai " is divided into subject " I ", predicate " liking " and object " Liang shanbo and Zhu yingtai " by analysis, and will not yet The object " Liang shanbo and Zhu yingtai " in the dictionary 2 is contained in as the second new new-added item, second adding set adds the second new new-added item Enter in the dictionary 2, and using piano classification as the second new new-added item " Liang shanbo and Zhu yingtai " classification.

Then, generating means 13 are described in detail with reference to the embodiment shown in FIG. 4, and are contained in by reference This, repeats no more.

One of preferred scheme as the present embodiment, the collator 1 also includes the 3rd acquisition device (being unillustrated) (it is unillustrated) with the first adding set.

Specifically, the 3rd acquisition device obtains the first new new-added item in the list entries, then, the first addition dress Put and add the described first new new-added item in the dictionary 2, and using the related category as the described first new new-added item classification.

Include " receiving for example, second acquisition device 14 obtains the application information related to user's input operation The identification information on part people column ", then first determining device 15 determine the application information related to user's input operation Related related category includes email address classification.When the user inputs list entries in " the addressee column "aaa@ bbb.com, then the 3rd acquisition device the first new new-added item is obtained in the list entriesaaa@bbb.com, then, first adds Feeder apparatus is by the first new new-added itemaaa@bbb.comAdd in the dictionary 2, and first increased newly email address classification as this The classification of item.

It is also understood by those skilled in the art that in embodiment shown in Fig. 6, performed by the first acquisition device 11 Operation has no sequencing with the operation performed by the second acquisition device 14, the first determining device 15 and the first adjusting apparatus 16, In fact, the first acquisition device 11 can first obtain the list entries from user, then, the second acquisition device 14, first is determined The adjusting apparatus 16 of device 15 and first sequentially works again；Or, the second acquisition device 14, the first determining device 15 and first are adjusted Engagement positions 16 first sequentially work, and then, the first acquisition device 11 obtains the list entries from user；Or, first obtains dress Put 11 and second acquisition device 14, the first determining device 15 and the first adjusting apparatus 16 perform operation etc. parallel.

, can be according to the current input environment of user according to the collator of the present invention, the weight of adjustment classification is believed in real time Breath so that the candidate item sequence of classification corresponding with the input environment that user is current is forward, so that the sequence of candidate item more meets The demand of user, is easy to user quickly to select correct candidate item, improves the input speed of user.

One of preferred scheme as the present invention, the collator 1 also includes the second adjusting apparatus (being unillustrated).Institute State the second adjusting apparatus and frequency is selected according to the history of the candidate item of each classification, to adjust the weight information of each classification.Its In, the mode for adjusting the weight information of each classification includes but is not limited to：

1) select the result that frequency is compared to each other to adjust the weight information of each classification based on each history.For example, described The history selection frequency that the acquisition of second adjusting apparatus belongs to computer other candidate item is 800 times, belongs to the time of music categories The history selection frequency of option is 500 times, and the history selection frequency for belonging to mechanical other candidate item is 10 times, computer classification Corresponding weight information is 2, and the corresponding weight information of music categories is 6, and the corresponding weight information of mechanical classification is 5, then described After more each history selection frequency of collator, determine that the history selection frequency of computer other candidate item is higher than music categories Candidate item history selection frequency, the history selection frequency of the candidate item of music categories is higher than mechanical other candidate item again History selects frequency, thus, and the corresponding weight information of computer classification is adjusted to be higher than music class by second adjusting apparatus Not corresponding weight information 7.

2) comparative result of frequency and predetermined threshold is selected to adjust the weight information of each classification based on history.For example, When second adjusting apparatus judges that the history selection frequency of the candidate item of a classification is higher than predetermined threshold, then described second adjust Engagement positions heighten the corresponding weight information of the category；In another example, when second adjusting apparatus judges the candidate of a classification The history selection frequency of item is less than predetermined threshold, then second adjusting apparatus turns down the corresponding weight information of the category.Its In, the amplitude heightened or turned down is predetermined amplitude.

It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment, Er Qie In the case of without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter From the point of view of which point, embodiment all should be regarded as exemplary, and be nonrestrictive, the scope of the present invention is by appended power Profit is required rather than described above is limited, it is intended that all in the implication and scope of the equivalency of claim by falling Change is included in the present invention.Any reference in claim should not be considered as to the claim involved by limitation.This Outside, it is clear that the word of " comprising " one is not excluded for other units or step, and odd number is not excluded for plural number.That is stated in system claims is multiple Unit or device can also be realized by a unit or device by software or hardware.The first, the second grade word is used for table Show title, and be not offered as any specific order.

Claims

1. a kind of method of the computer implemented candidate item sequence for being used to determine input method, wherein, this method includes following step Suddenly：

A obtains the list entries from user；

B is inquired about according to the list entries in dictionary, to obtain the multiple candidates matched with the list entries , and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item performed after adjustment operation, come Processing is ranked up to the multiple candidate item, to obtain the ranking results of the sequence processing, wherein, the adjustment operation base Performed in current application environmental information；

C shows information to be supplied to the user according to the ranking results, generation；

Wherein, this method is further comprising the steps of：

- obtain the current application environmental information related to the input operation performed by the user；

- determine the related category related to the current application environmental information；

- the adjustment operation for adjusting the corresponding weight information of the related category is performed according to the current application environmental information.

2. according to the method described in claim 1, wherein, the step b is further comprising the steps of：

- according to the list entries, inquired about in the dictionary, to obtain the multiple times matched with the list entries Option, and the corresponding weight information of classification and the related letter of other sequences according to belonging to each candidate item in the plurality of candidate item Breath, to be ranked up processing to the multiple candidate item, to obtain the ranking results of the sequence processing.

3. method according to claim 2, wherein, other described sequence relevant informations include at least one of following：

The corresponding word frequency information of each candidate item in-the multiple candidate item；

Each matching degree between candidate item and the list entries in-the multiple candidate item.

4. according to the method described in claim 1, wherein, this method is further comprising the steps of：

- the first new new-added item is obtained in the list entries；

- the described first new new-added item is added in the dictionary, and using the related category as the described first new new-added item classification.

5. according to the method described in claim 1, wherein, the current application environmental information includes at least one of following：

- the application information related to the input operation；

- program the scene information related to the input operation；

The text message included in-application scenarios related to the input operation.

6. method according to claim 5, wherein, the current application environmental information includes the text message, its In, this method is further comprising the steps of：

- the second new new-added item is determined according to the text message；

- the described second new new-added item is added in the dictionary, and using the related category as the described second new new-added item classification.

7. method according to any one of claim 1 to 6, wherein, the step b is further comprising the steps of：

- according to the list entries, in the dictionary and the application program related to the input operation performed by the user Inquired about in the program dictionary of offer, to obtain the multiple candidate items matched with the list entries, and according to the plurality of The corresponding weight information of classification in candidate item belonging to each candidate item, to be ranked up processing to the multiple candidate item, with Obtain the ranking results of the sequence processing.

8. method according to claim 7, wherein, this method is further comprising the steps of：

- add the user selected candidate item from described program dictionary in the dictionary.

9. according to the method described in claim 1, wherein, this method is further comprising the steps of：

- frequency is selected according to the history of the candidate item of each classification, to adjust the weight information of each classification.

10. according to the method described in claim 1, wherein, this method is further comprising the steps of：

- obtain the renewal request that request updates candidate item that is used for from the user；

- according to it is described update request and it is described show information, candidate item to be showed is selected in the ranking results；

- according to the candidate item to be showed, update described in show information.

11. according to the method described in claim 1, wherein, this method is further comprising the steps of：

Wherein, this method is further comprising the steps of：

- judge when according to the renewal request and the information that shows in the ranking results not comprising the time for needing to be showed During option, the classification according to belonging to the candidate item being each not included in the ranking results in the multiple candidate item is corresponding Weight information, to be ranked up processing to the candidate item being not included in the ranking results, with the row after being updated Sequence result；

- according to the ranking results after the renewal, update described in show information to be supplied to the user.

12. a kind of collator for the candidate item sequence for being used to determine input method, wherein, the collator includes：

First acquisition device, for obtaining the list entries from user；

Sort processing unit, for according to the list entries, being inquired about in dictionary, with acquisition and the list entries phase Multiple candidate items of matching, and the classification correspondence according to belonging to each candidate item in the plurality of candidate item performed after adjustment operation Weight information, to be ranked up processing to the multiple candidate item, with obtain it is described sequence processing ranking results, wherein, The adjustment operation is performed based on current application environmental information；

Generating means, for according to the ranking results, generation to show information to be supplied to the user；

Wherein, the collator also includes：

Second acquisition device, the current application environmental information related to the input operation performed by the user for acquisition；

First determining device, the related category related to the current application environmental information for determination；

First adjusting apparatus, for being performed according to the current application environmental information adjust the corresponding weight of the related category and believe The adjustment operation of breath.

13. collator according to claim 12, wherein, the sequence processing unit also includes：

First sub- collator, for according to the list entries, being inquired about in the dictionary, with obtain with the input Multiple candidate items that sequence matches, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item And other sequence relevant informations, to be ranked up processing to the multiple candidate item, to obtain the sequence of the sequence processing As a result.

14. collator according to claim 13, wherein, other described sequence relevant informations include following at least one ：

15. collator according to claim 13, wherein, the collator also includes：

3rd acquisition device, in the list entries obtain the first new new-added item；

First adding set, for the described first new new-added item to be added in the dictionary, and using the related category as described The classification of first new new-added item.

16. collator according to claim 13, wherein, the current application environmental information includes following at least one ：

- the application information related to the input operation；

- program the scene information related to the input operation；

17. collator according to claim 16, wherein, the current application environmental information includes the text envelope Breath, wherein, the collator also includes：

Second determining device, for determining the second new new-added item according to the text message；

Second adding set, for the described second new new-added item to be added in the dictionary, and using the related category as described The classification of second new new-added item.

18. the collator according to any one of claim 13 to 17, wherein, the sequence processing unit also includes：

3rd sub- collator, for according to the list entries, the dictionary and with the input performed by the user Inquired about in the program dictionary that the related application program of operation is provided, to obtain the multiple times matched with the list entries Option, and the corresponding weight information of classification according to belonging to each candidate item in the plurality of candidate item, to the multiple candidate Item is ranked up processing, to obtain the ranking results of the sequence processing.

19. collator according to claim 18, wherein, the collator also includes：

3rd adding set, for the selected candidate item from described program dictionary of the user to be added into the dictionary In.

20. collator according to claim 12, wherein, the collator also includes：

Second adjusting apparatus, the history for the candidate item according to each classification select frequency, to adjust the weight of each classification Information.

21. collator according to claim 12, wherein, the collator also includes：

4th acquisition device, for obtain from the user be used for ask update candidate item renewal request；

Selection device, for according to it is described update request and it is described show information, select to wait to show in the ranking results Candidate item；

First updating device, for showing information described according to the candidate item to be showed, updating.

22. collator according to claim 12, wherein, the collator also includes：

5th acquisition device, for obtain from the user be used for ask update candidate item renewal request；

Sub-processing unit, for when according to it is described update request and the information that shows judge not include in the ranking results When needing the candidate item showed, according to the candidate item institute being each not included in the multiple candidate item in the ranking results The corresponding weight information of classification of category, to be ranked up processing to the candidate item being not included in the ranking results, with Ranking results after being updated；

Second updating device, for showing information to be supplied to the use described according to the ranking results after the renewal, updating Family.

23. a kind of computer equipment, wherein, the computer equipment includes the row as described at least one in claim 12 to 22 Sequence device.