WO2012139475A1 - 获取与输入按键序列相对应的候选字符串的方法与设备 - Google Patents

获取与输入按键序列相对应的候选字符串的方法与设备 Download PDF

Info

Publication number
WO2012139475A1
WO2012139475A1 PCT/CN2012/073589 CN2012073589W WO2012139475A1 WO 2012139475 A1 WO2012139475 A1 WO 2012139475A1 CN 2012073589 W CN2012073589 W CN 2012073589W WO 2012139475 A1 WO2012139475 A1 WO 2012139475A1
Authority
WO
WIPO (PCT)
Prior art keywords
candidate
user
input
key sequence
input key
Prior art date
Application number
PCT/CN2012/073589
Other languages
English (en)
French (fr)
Inventor
秦添
Original Assignee
百度在线网络技术（北京）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 百度在线网络技术（北京）有限公司 filed Critical 百度在线网络技术（北京）有限公司
Publication of WO2012139475A1 publication Critical patent/WO2012139475A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/02—Input arrangements using manually operated switches, e.g. using keyboards or dials
    • G06F3/023—Arrangements for converting discrete items of information into a coded form, e.g. arrangements for interpreting keyboard generated codes as alphanumeric codes, operand codes or instruction codes
    • G06F3/0233—Character input methods
    • G06F3/0237—Character input methods using prediction or retrieval techniques

Definitions

  • the present invention relates to the field of input method technologies, and in particular, to a technique for acquiring a candidate character string corresponding to an input key sequence. Background technique
  • the existing input method usually performs a query in the thesaurus based on the input sequence of the user input, and obtains the corresponding matching term. Further, the existing input method combines the user's language habits and the like to establish a fuzzy sound rule, such as mapping "Z" to
  • the existing input method fails to take into account input errors due to factors such as the user's habit of tapping the keyboard, and an extension of the input key sequence based on the keyboard layout.
  • the user often mistakes the adjacent keys of the desired keys when making the input, so that the user equipment has difficulty obtaining the character string that the user wishes to input, so that the process of querying the thesaurus based on the string to obtain the candidate terms is even more difficult to talk about. Accordingly, this also greatly limits the user's input speed and input accuracy, while reducing user input efficiency and user experience.
  • a method for obtaining a candidate character string corresponding to an input key sequence is provided, wherein the method comprises the following steps:
  • an apparatus for obtaining a candidate character string corresponding to an input key sequence comprising:
  • An input obtaining device configured to acquire an input button sequence input by a user via a keyboard of the user equipment
  • the string obtaining means is configured to perform a query in the candidate lexicon according to the input key sequence and the keyboard layout of the keyboard to obtain one or more candidate character strings corresponding to the input key sequence.
  • the present invention is based on a user input input key sequence and a keyboard layout of a user's input keyboard, such as a QWERTY full keyboard, a normal phone keypad, an alphabetical full keyboard, and any customization.
  • the keyboard layout of the layout keyboard, etc. is queried in the candidate lexicon to obtain the corresponding candidate character string, thereby realizing the expansion of the input key sequence input by the user based on the key on the keyboard, and obtaining corresponding corresponding Get candidate strings for candidate terms.
  • the invention reduces the input error caused by the user's habit of tapping the keyboard and the like, and adds an extension of the input key sequence based on the keyboard layout, thereby effectively improving the input efficiency of the user and improving the user experience.
  • 1 shows a schematic diagram of an apparatus for acquiring candidate character strings corresponding to an input key sequence, in accordance with an aspect of the present invention
  • 2 shows a schematic diagram of an apparatus for acquiring candidate character strings corresponding to an input key sequence in accordance with a preferred embodiment of the present invention
  • FIG. 3 is a diagram showing an apparatus for acquiring a candidate character string corresponding to an input key sequence according to another preferred embodiment of the present invention.
  • FIG. 4 shows a flow chart of a method for acquiring a candidate character string corresponding to an input key sequence in accordance with another aspect of the present invention
  • Figure 5 is a flow chart of a method for obtaining a candidate character string corresponding to an input key sequence in accordance with a preferred embodiment of the present invention
  • FIG. 6 is a flow chart of a method for obtaining a candidate character string corresponding to an input key sequence in accordance with another preferred embodiment of the present invention.
  • the user equipment 1 includes an input obtaining device 11 and a character string obtaining device 12. Specifically, the input obtaining device 11 acquires an input key sequence input by the user via the keyboard of the user equipment 1; the character string obtaining device 12 performs a query in the candidate lexicon according to the input key sequence and the keyboard layout of the keyboard, Obtaining one or more candidate character strings corresponding to the input key sequence.
  • the user equipment 1 can be any electronic product capable of interacting with a user through a keyboard, such as a computer, a mobile phone, a PDA, a tablet, a PSP, a PPC, etc.; wherein the keyboard can be any physical type Keyboard or virtual keyboard, including but not limited to QWERTY full keyboard, normal phone keyboard (keypad), alphabetical full keyboard, and any custom layout virtual keyboard.
  • a keyboard such as a computer, a mobile phone, a PDA, a tablet, a PSP, a PPC, etc.
  • the keyboard can be any physical type Keyboard or virtual keyboard, including but not limited to QWERTY full keyboard, normal phone keyboard (keypad), alphabetical full keyboard, and any custom layout virtual keyboard.
  • the input acquisition means 11 acquires a keyboard input by the user via the user device 1 Input key sequence.
  • the input obtaining device 11 obtains an input key sequence input by the user via the keyboard of the user device 1 by detecting and parsing an electric signal generated by the user pressing a virtual keyboard on the touch screen of the user device 1 or tapping a physical keyboard or the like.
  • the user equipment 1 is an electronic product with a touch screen
  • the corresponding virtual keyboard is a QWERTY full keyboard.
  • the user inputs the user equipment 1 by pressing a button in the QWERTY full keyboard, and accordingly, the input obtaining device 11 passes The electrical signal generated by the user pressing the touch screen is detected and analyzed to obtain an input key sequence input by the user.
  • the string obtaining means 12 performs a query in the candidate lexicon according to the input key sequence in combination with the keyboard layout of the keyboard to acquire one or more candidate character strings corresponding to the input key sequence. Specifically, the character string obtaining device 12 performs a query in the candidate lexicon according to the input key sequence acquired by the input obtaining device 11 and the keyboard layout of the user equipment 1 to acquire one or more candidates corresponding to the input key sequence. String.
  • the manner in which the character string obtaining device 12 acquires the candidate character string includes, but is not limited to: 1) a character mapping table based on a preset keyboard layout, such as "g ⁇ f, h, t, v", “s ⁇ a , d, w, x”, etc., obtain one or more mapping strings corresponding to the input key sequence, such as obtaining the mapping strings “hongs”, “fongs”, “honha”, “tonfd”, etc.
  • the keys corresponding to the first extended character strings are closer to the input key sequence, such as obtaining the first extended character string “sun” according to the input key sequence “sin” , “san” and “son”, according to the keyboard layout, select candidate strings from these first extended strings, such as “u”, “0” and “ ⁇ adjacent to the button position, then select "sun” and “"son” as a candidate string, here, "distance” refers to the first expansion
  • the character string obtaining means 12 obtains corresponding candidate character strings according to the above-mentioned three types of obtaining candidate character strings according to the input key sequence of the user, and merges the candidate character strings obtained by the three methods to obtain the final user.
  • the input key sequence corresponds to the candidate string.
  • the candidate vocabulary includes a local vocabulary and a network vocabulary; and, the candidate vocabulary further includes a Chinese lexicon, an English lexicon, other language lexicons, and a mixed vocabulary of multiple languages.
  • the input obtaining means 11 and the character string obtaining means 12 are continuously operated.
  • the input obtaining device 11 is configured to acquire an input key sequence input by the user via the keyboard of the user equipment 1;
  • the string obtaining device 12 is configured to perform a query in the candidate lexicon according to the input key sequence and the keyboard layout. And acquiring one or more candidate character strings corresponding to the input key sequence.
  • continuous means that the input obtaining means 11 and the character string obtaining means 12 respectively perform the acquisition of the input key sequence and the acquisition of the candidate character string according to the set or real-time adjusted working mode requirements. Until the user stops inputting the input key sequence via the keyboard of the user device 1 for a long time.
  • the string obtaining device 12 includes a first expanding unit 121 and a selecting unit 122.
  • the first expansion unit 121 performs a query in the candidate vocabulary according to the input key sequence to obtain the One or more first extended character strings corresponding to the key sequence;
  • the selecting unit 122 selects the one or more candidate character strings from the one or more first extended character strings according to the keyboard layout.
  • the first expansion unit 121 performs a query in the candidate vocabulary according to the input key sequence to acquire one or more first extended character strings corresponding to the input key sequence. Specifically, the first expansion unit 121 performs a query in the candidate lexicon according to the input key sequence input by the user via the keyboard of the user equipment 1 to acquire one or more first extended strings corresponding to the input key sequence. . For example, assuming that the key sequence "sin" is input, the candidate vocabulary is the local lexicon of the user equipment 1, and the first expansion unit 121 performs an extended query on the "sin” in the candidate lexicon to obtain a corresponding first extended string. "sun", “san” and "son”.
  • first extended character string is only an example, and other existing or future possible ways of obtaining the first extended character string may be applicable to the present invention, and should also be included in the present invention. Within the scope of protection, and is hereby incorporated by reference.
  • the selecting unit 122 selects the one or more candidate character strings from the one or more first expanded character strings according to the keyboard layout. Specifically, the selecting unit 122 selects one or more candidate character strings corresponding to the keyboard layout from the one or more first extended character strings acquired by the first expanding unit 121 according to the keyboard layout of the user input keyboard. For example, in the above example, the first expansion unit 121 obtains the first extended strings “sun”, “san”, and “son”, and the user's input keyboard is a QWERTY full keyboard, "u”, "o", and “i”. The key positions are adjacent to each other, and the selection unit 122 selects "sun” and "son” as the candidate character strings from these first extended character strings. Those skilled in the art should understand that the manner of obtaining the candidate character string is only an example. Other existing or future possible methods for obtaining the candidate character string may be applicable to the present invention and should also be included in the scope of the present invention. And is included here by reference.
  • FIG. 3 is a schematic diagram of a device showing a user device that obtains a candidate character string corresponding to an input key sequence, in accordance with another preferred embodiment of the present invention.
  • the string obtaining device 12" includes a second expanding unit 123" and a query unit 124". Specifically, the second expanding unit 123" acquires the input button sequence according to the input button sequence in combination with the keyboard layout. Corresponding one or more second extended strings; the query unit 124" is in the candidate lexicon Performing a matching query on the one or more second extended strings to obtain the one or more candidate strings.
  • the second extension unit 123" of the character string obtaining means 12" acquires one or more second extended character strings corresponding to the input key sequence in accordance with the input key sequence in conjunction with the keyboard layout. Specifically, the second extension unit 123" acquires one or more second extended character strings corresponding to the input key sequence in accordance with the input key sequence input by the user via the user device 1, in conjunction with the keyboard layout of the user device 1. For example. Assume that the input key sequence "uing", the keyboard of the user device 1 is a QWERTY full keyboard, according to the keyboard layout, "u” is adjacent to the "y", "i", "j” button positions, according to which, the second extension Unit 123" obtains the second extended string “ying", "iing", "jing".
  • the user device 1 is a mobile phone
  • the keyboard is a normal mobile phone keyboard, that is, a keyboard corresponding to a plurality of letters, according to the keyboard layout, the button "ghi", the button “mno” and the button
  • the “jkl” button positions are adjacent, and accordingly, the second extension unit 123" obtains the second extension strings “gent”, “hent”, “ient”, “ment”, “nent”, “oent”, “kent” , "lent".
  • the manner of obtaining the second extended string is only an example, and other existing or future possible ways of obtaining the second extended string may be applied to the present invention. It is intended to be included within the scope of the invention and is hereby incorporated by reference.
  • the query unit 124" performs a matching query on the one or more second extended strings in the candidate vocabulary to obtain the one or more candidate strings. Specifically, the query unit 124" Performing a matching query on one or more second extended character strings obtained from the second extension unit 123" in the candidate vocabulary to obtain one or more candidate character strings. For example, in the above example, the second extension unit 123" The second extended strings “ying”, “iing", and “jing” are obtained, and the query unit 124" performs a matching query on the second extended character strings in the candidate thesaurus to obtain the candidate strings "ymg" and "jmg".
  • the second extension unit 123 obtains the second extended character string "gent", “hent”, “ient”, “ment”, “nent”, “oent”, “kent”, “lent”, and the query unit 124 "Complete the query for these second extended strings in the candidate lexicon to obtain the candidate strings "gent", “hent”, “kent”, “lent”.
  • the user equipment 1 further comprises screening means (not shown) for filtering the one or more candidate strings according to the query assistance information to obtain One or more candidate strings after filtering.
  • screening means not shown
  • the input key acquisition device 12 acquires one or more candidate characters according to the input key sequence, based on the input key sequence input by the user via the user equipment.
  • the specific process of the string is the same as that performed by the input obtaining means 11 and the character string obtaining means 12 in the embodiment described above with reference to Fig. 1. For the sake of brevity, it is hereby incorporated by reference.
  • the screening device filters one or more candidate character strings obtained by the string obtaining device 12 according to the query auxiliary information, such as the current application of the user equipment 1, the user characteristic information of the user, or the like, or
  • the obtaining means 12 combines the acquisition and screening of the candidate string to obtain one or more candidate strings after the screening.
  • the candidate character string obtained by the string obtaining device 12 according to the input key sequence "sin” is "son” and "sun”
  • the current application of the user device 1 is a contact input of the mobile phone phonebook, and is input in Chinese
  • the current application filters the candidate strings "son” and "sun” to obtain the filtered candidate string "sun”.
  • the input key sequence is "sin”
  • the user's input keyboard is a QWERTY full keyboard
  • the filtering device is combined with the character string obtaining device 12, according to the user's previous input history "son”
  • the input key sequence "sin” "And the keyboard layout of the input keyboard, such as "i” adjacent keys are "u” and "0”, query in the candidate lexicon, get the candidate string "son”.
  • the manner of filtering the candidate character string according to the query auxiliary information is only an example, and other existing or future possible methods for screening candidate character strings according to the query auxiliary information may be applicable to the present invention, and It is intended to be included within the scope of the invention and is hereby incorporated by reference.
  • the query assistance information includes, but is not limited to, at least one of the following:
  • the current application of the user equipment 1 refers to an application currently running by the user equipment 1.
  • the screening device filters one or more candidate character strings acquired by the string obtaining device 12 according to the application currently running by the user device 1, or the filtering device is combined with the string obtaining device 12 according to the application currently running by the user device 1.
  • the input key sequence input by the user and the keyboard layout of the input keyboard perform matching query in the candidate lexicon, and filter the candidate character string so that the candidate character string not only satisfies the input key sequence based on the keyboard layout. Expanded, and conforms to the current application of the user device 1.
  • the input key sequence input by the user is "sin”
  • the current application of the user device 1 is the contact input of the phone book
  • the character string obtaining means 12 acquires the corresponding candidate character string as "son”, "sun”
  • the screening device according to The current application sorts these candidate strings to obtain the candidate string "sun”.
  • the user feature information of the user includes the user's input history record, the user's input preference setting, and the like, and the screening device selects the character string obtaining device 12 according to the user's user feature information, such as the user's input history record, the user's input preference setting, and the like.
  • the obtained one or more candidate strings are filtered, or the filtering device is combined with the string obtaining device 12, and is performed in the candidate lexicon according to the user characteristic information of the user, and the input key sequence input by the user and the keyboard layout of the input keyboard. Matching the query, filtering obtains the candidate string, so that the candidate string(s) not only satisfy the extension of the input key sequence based on the keyboard layout, but also conforms to the user characteristic information of the user.
  • the user's input preference is set to "m ⁇ n", that is, the user habitually hits “m” when tapping "n”, assuming the input key sequence is “ming”, and the input keyboard is arranged in alphabetical order.
  • the full keyboard, the filtering device is combined with the string obtaining device 12, according to the input preference setting, and the input key sequence and the keyboard layout, such as "m” adjacent keys are “n” and “ ⁇ , determine the right” Ming "The expanded string is "ning”, and the "ning" is queried in the candidate lexicon to obtain the candidate string "ning".
  • the prior input information of the input key sequence includes the record recorded in the cache of the user device 1
  • the character string and the term input before the input key sequence, and the screening device analyzes the context semantics between the candidate input string(s) and the candidate character string obtained by the string obtaining device 12, and the candidate string(s) Performing a screening, or combining the filtering device with the string obtaining device 12, performing matching query and context semantic analysis in the candidate lexicon according to the prior input information(s), and the input key sequence input by the user and the keyboard layout of the input keyboard
  • the candidate string is obtained by filtering so that the candidate string(s) not only satisfy the expansion of the input key sequence based on the keyboard layout, but also conform to the semantics of the prior input information.
  • the input key sequence input by the user is “ahiwuduo”
  • the string obtaining means 12 acquires the corresponding candidate character string, such as "shiwusuo", “zhisuduo”, “shiwuduo”, “zhiwusuo”, which is recorded in the cache of the user device 1.
  • the entry entered by the user before inputting the input key sequence is "lawyer", and the screening device performs context semantic analysis on the prior input information and the candidate character string respectively, and determines that the candidate character string is "shiwusuo".
  • the semantic analysis result corresponding to the candidate character string includes a result obtained by semantic analysis of the candidate character string itself, such as whether the candidate character string is meaningful, and the filtering means performs semantic analysis on the candidate character string obtained by the string obtaining device 12
  • the candidate character string is filtered according to the semantic analysis result(s), or the filtering device is combined with the string obtaining device 12, and the matching query is performed in the candidate lexicon according to the input key sequence input by the user and the keyboard layout of the input keyboard.
  • the candidate string is semantically analyzed, and according to the result of the semantic analysis, the candidate string is obtained by screening, so that the candidate string(s) not only satisfy the expansion of the input key sequence based on the keyboard layout, but also the candidate string itself is meaningful. .
  • the input keyboard of the user is a QWERTY full keyboard
  • the character string obtaining means 12 is based on the input key sequence and the keyboard layout, such as "the adjacent key of the keyboard is “o” , “u”, etc.
  • the adjacent keys of "s” are “d”, “w”, “x”, etc., and are queried in the candidate lexicon to obtain candidate strings “some people”, “dime people” and " Dome people”
  • the screening device separately performs semantic analysis on these candidate strings, and determines candidate strings according to the corresponding semantic analysis results, such as "some people”
  • the distance threshold of the candidate string and the input key sequence includes a preset maximum value of the number of characters between the candidate character string and the input key sequence, and the characters in the candidate string are lost.
  • the screening device filters the candidate character string acquired by the string obtaining device 12 according to the distance threshold, or the filtering device is combined with the string obtaining device 12, according to the distance threshold, and the input key sequence input by the user and the keyboard layout of the input keyboard. Perform matching query in the candidate lexicon, and filter the candidate string so that the candidate string(s) not only satisfy the extension of the input key sequence based on the keyboard layout, but also does not exceed the candidate character string and the input key sequence.
  • the user input key sequence is "ahurufa", assuming that the preset distance threshold is 1, where the distance threshold refers to a preset maximum value of the number of characters between the candidate character string and the input key sequence, that is, the candidate string is at most only It can be different from the input key sequence by one character.
  • the user's input keyboard is a QWERTY full keyboard, and the filtering device is combined with the character string obtaining device 12, according to the distance threshold, and the input key sequence input by the user and the keyboard layout of the input keyboard.
  • the adjacent keys of "a” are “z”, “s”, etc.
  • the adjacent keys of "u” are “ ⁇ , "y”, etc., and are queried in the candidate lexicon, and the candidate string is selected as "Shurufs","zhurufa”.
  • the user input key sequence is "jind”, assuming that the distance threshold is 1, where the distance threshold refers to the corresponding key position of the character in the input keyboard and the input key sequence in the candidate string.
  • the input keyboard is a QWERTY full keyboard
  • the screening device is combined with the string obtaining device 12, according to the distance threshold, and the user Enter the input key sequence and the keyboard layout of the input keyboard, determine that the adjacent keys of "j" are “k”, “m”, “h”, etc., the adjacent keys of "i” are “u”, etc., and are candidates
  • the query is performed in the thesaurus to obtain candidate strings “kind”, “mind”, “hind”, "jund”.
  • the above-mentioned query auxiliary information can be used not only for screening candidate strings, but also for combining candidate pairs or even combining them for screening candidate strings.
  • the above-mentioned query auxiliary information is only an example, and other existing or future query auxiliary information may be applicable to the present invention, and should also be included in the scope of the present invention, and The way is included here.
  • the user characteristic information includes, but is not limited to, at least one of the following:
  • the input history of the user including but not limited to the input key sequence of the user history input, the user's selection operation of the candidate character string, the mapping relationship between the input key sequence of the user history input, and the candidate string selected by the user, etc. If the input key sequence is "sm” and the corresponding candidate strings are “sun” and “son”, the user selects the candidate character string “sun”, and the user device 1 establishes a mapping relationship between "sin” and "sun”.
  • the user's input preference settings including but not limited to the mapping relationship between the characters specially set by the user according to their input habits, such as "m ⁇ n", and the user's preference for the input method, such as the high priority of the English input. The priority entered in Chinese.
  • the above user characteristic information can be used not only for screening candidate strings, but also for combining two or two, or even combining them for filtering candidate strings.
  • the above-mentioned user characteristic information is only an example, and other existing or future user feature information may be applicable to the present invention, and should also be included in the scope of protection of the present invention. Included here.
  • the user equipment 1 further comprises a character string providing means (not shown) for providing said one or more candidate character strings to the user equipment 1.
  • a character string providing means (not shown) for providing said one or more candidate character strings to the user equipment 1.
  • the other preferred embodiment is described in detail below with reference to FIG. 1, wherein the input key acquisition device 12 acquires one or more input key sequences according to the input key sequence.
  • the candidate string is the same as the process performed by the input obtaining means 11 and the character string obtaining means 12 in the embodiment described above with reference to FIG. 1. For the sake of brevity, it is incorporated herein by reference. .
  • the string providing means supplies the one or more candidate character strings acquired by the string obtaining means 12 to the random or in a certain order, such as the word frequency of the candidate character string from high to low, the order of the candidates in the candidate thesaurus, etc.
  • User equipment 1 For example, the candidate character strings acquired by the string obtaining means 12 are "son” and “sun", and the word frequency of the candidate character strings in the candidate lexicon is used, for example, the word frequency of "sun” is higher than the word frequency of "son”.
  • the string providing means supplies the candidate strings to the user equipment 1 in the order of the word frequency from high to low, namely "sun", "son”.
  • the string providing device will also be able to combine the screening device with the string obtaining device 12
  • One or more candidate strings are taken to the user equipment 1 in the manner described above.
  • the manner of providing the candidate character string is only an example, and other existing or future possible ways of providing the candidate character string may be applicable to the present invention and should also be included in the scope of the present invention. And is included here by reference.
  • the character string providing means provides the candidate character string to the user equipment 1.
  • the user can select the candidate character string that he/she wishes to input through the user equipment 1 for acquiring the corresponding candidate term;
  • the step can also implement an extended reminder of the input key sequence input by the user.
  • the candidate character string is an input error reminder for the input key sequence, which can enable the user to understand the wrong input habit in time.
  • the user equipment 1 further includes a first updating means (not shown) for updating the candidate vocabulary based on the user's selection operation of the one or more candidate character strings.
  • the first update device updates the candidate vocabulary according to the selection operation of the one or more candidate character strings provided by the user to the character string providing device, such as updating the word frequency of the candidate character string selected by the user, updating the input key sequence and the candidate The mapping relationship of strings, etc.
  • the user performs a selection operation on the candidate character strings via a keyboard or a mouse, such as the user selecting the candidate character string "son”
  • the first update means updates the word frequency of "son” in the candidate lexicon according to the selection operation of the user, that is, the word "son” is incremented by one.
  • the input key sequence is "sin”
  • the corresponding candidate character strings provided by the character string providing device to the user equipment 1 are “son” and "sun”
  • the user selects the candidate character string "son”
  • the first update device selects the candidate string according to the selection.
  • the user equipment 1 further includes a term acquisition means (not shown) and a term providing means (not shown), and the term acquisition means is based on the one or more candidates a string, performing a matching query in the matching thesaurus to obtain one or more candidate terms corresponding to the one or more candidate character strings; the term providing means One or more candidate terms are provided to the user device 1.
  • a further preferred embodiment of the present invention is described in detail below with reference to FIG. 1.
  • the input character acquisition device 12 acquires one or more input key sequences according to the input key sequence.
  • the candidate string has the same specific procedure as that performed by the input obtaining means 11 and the character string obtaining means 12 in the embodiment described above with reference to FIG. 1, and is included herein by way of citation for the sake of brevity. .
  • the vocabulary obtaining means performs a matching query in the matching vocabulary based on one or more candidate character strings obtained by the character string obtaining means 12 to obtain one or more candidate terms corresponding to the candidate character string(s). For example, the string obtaining means 12 obtains the candidate character strings "sun”, "son”, and the term obtaining means performs a matching query on the candidate character strings in the matching thesaurus to obtain corresponding candidate terms "sun” and "son”. .
  • the term acquiring device may further perform a matching query on the candidate character string corresponding to the user's selection operation in the matching vocabulary to obtain a corresponding candidate term; assuming that the user selects the candidate character string "sun",
  • the term acquisition device only performs a matching query on the candidate string "sun” in the matching thesaurus to obtain the corresponding candidate term "sun".
  • the matching vocabulary can be independent of the lexical vocabulary or can be combined into one vocabulary.
  • the term providing means provides the user equipment 1 with one or more candidate terms acquired by the term acquiring means.
  • the term acquisition device acquires the candidate terms "sun” and "son”, according to the frequency of use of the candidate terms in the candidate lexicon, such as “sun” users use frequencies higher than “son” Frequency, the term providing means provides these candidate terms to the user equipment 1 in descending order of user frequency, that is, "sun” and "son”.
  • the above manner of providing candidate terms is only an example, and other existing or future possible ways of providing candidate terms may be applicable to the present invention and should also be included in the scope of the present invention. And is included here by reference.
  • the term providing means provides one or more candidate terms acquired by the term acquiring means to the user equipment 1 on the one hand, for the user to select a candidate term that he wishes to input; on the other hand, the step is also implemented.
  • the extension of the user-selectable candidate term further reduces the error correction process when the user inputs an error, and also makes the accuracy of the user input become Possible.
  • the user equipment 1 further includes matching means (not shown) for performing a matching query in the matching vocabulary according to the input key sequence to obtain a sequence corresponding to the input key sequence.
  • One or more matching terms the term providing means provides the one or more matching terms to the user device 1 with the one or more candidate terms.
  • the matching device performs a matching query in the matching vocabulary according to the input key sequence acquired by the input obtaining device 11 to obtain one or more matching terms corresponding to the input key sequence; subsequently, the vocabulary providing device
  • the one or more matching terms acquired by the matching device and the candidate term acquired by the term acquiring device are also provided to the user equipment 1, such as combining the matching term and the candidate term, and further, the word frequency may be selected to be the highest.
  • a plurality of matching terms and a plurality of candidate terms with the highest word frequency are combined, and are provided to the user equipment 1 according to the word frequency of the combined words, the frequency of use by the user, and the like.
  • the candidate lexicon and the matching vocabulary are a thesaurus
  • the matching device performs a matching query on the input key sequence in the lexicon, and obtains a corresponding matching term "thinking”
  • the string obtaining means 12 obtains corresponding candidate strings, such as “son” and “sun”
  • the entry obtaining means obtains the corresponding candidate terms "son", "sun” "
  • the term providing means provides the word terms and the candidate terms in the order of the word frequency from high to low according to the word frequency of the matching terms and the candidate terms in the thesaurus, such as "thinking” "Sun”, “four years”, “raging”, “son”, “private thoughts”.
  • the term providing means provides the matching term and the candidate term to the user device 1, on the one hand, providing the user with more choices; on the other hand, in the case where the matching term is not the term that the user wishes to input
  • the candidate term is provided, which reduces the number of times the user needs to re-enter, and improves the input efficiency and input accuracy of the user.
  • the user equipment 1 further includes a selection obtaining means (not shown) for obtaining a selection operation of the one or more candidate character strings by the user; And performing, according to the candidate character string corresponding to the selecting operation, performing a matching query in the matching vocabulary to obtain one or more candidate terms corresponding to the candidate character string corresponding to the selecting operation.
  • the selection obtaining device acquires a selection operation of the user for one or more candidate character strings; the vocabulary obtaining device further performs a matching query in the matching vocabulary according to the candidate character string corresponding to the selection operation, to obtain a One or more candidate terms corresponding to the candidate character string corresponding to the selection operation.
  • the selection obtaining means acquires the user's selection operation on the candidate character strings, such as the user selecting the candidate character string "sun” through the keyboard; the entry obtaining means according to "sun” Matching queries are made in the matching lexicon, and candidate terms “sun”, “loss”, “bamboo”, etc. are obtained.
  • Matching queries are made in the matching lexicon, and candidate terms “sun”, “loss”, “bamboo”, etc. are obtained.
  • the user equipment 1 further includes a second updating means (not shown) that updates the matching vocabulary based on the user's selection operation of the one or more candidate terms.
  • the second update device updates the matching vocabulary according to the user's selection operation on the candidate term, such as updating the word frequency of the candidate term selected by the user in the matching vocabulary, updating the input key sequence, and the candidate term selected by the user. Mapping relationships, etc. For example, assuming that the key sequence "sin" is input, the character string obtaining means 12 acquires the corresponding candidate characters "son” and "sun”, and the entry obtaining means acquires the corresponding candidate terms "sun” and "son", and the entry is provided.
  • the device provides the two candidate terms to the user equipment 1 , and the user selects the candidate term “sun” via the user device 1 , and the second update device updates the word frequency of the “sun” in the matching the thesaurus according to the selection operation, that is, The word "sun” is incremented by one, and the mapping relationship between the input key sequence "sin” and the candidate term “sun” selected by the user is updated.
  • the manner of updating the matching lexicon is only an example, and other existing or future possible ways of updating the matching lexicon, as applicable to the present invention, are also included in the scope of the present invention. And is included here by reference.
  • step S1 the user equipment 1 acquires an input key sequence input by the user via the keyboard of the user equipment 1; subsequently, in step S2, the user equipment 1 combines the keyboard layout of the keyboard according to the key sequence, A query is made in the candidate lexicon to obtain one or more candidate character strings corresponding to the input key sequence.
  • the user equipment 1 can be any electronic product capable of interacting with a user through a keyboard, such as a computer, a mobile phone, a PDA, a tablet, a PSP, a PPC, etc.; wherein the keyboard can be any physical type Keyboard or virtual keyboard, including but not limited to QWERTY full keyboard, normal phone keypad, alphabetical full keyboard, and virtual keyboard for any custom layout.
  • a keyboard such as a computer, a mobile phone, a PDA, a tablet, a PSP, a PPC, etc.
  • the keyboard can be any physical type Keyboard or virtual keyboard, including but not limited to QWERTY full keyboard, normal phone keypad, alphabetical full keyboard, and virtual keyboard for any custom layout.
  • step S1 the user device 1 acquires an input key sequence input by the user via the keyboard of the user device 1.
  • the user equipment 1 obtains an input of the user input through the keyboard of the user equipment 1 by detecting and parsing an electrical signal generated by the user pressing a virtual keyboard on the touch screen of the user equipment 1 or tapping a physical keyboard or the like. Key sequence.
  • the user equipment 1 is an electronic product with a touch screen, and the corresponding virtual keyboard is a QWERTY full keyboard.
  • the user inputs the user equipment 1 by pressing a button in the QWERTY full keyboard, and accordingly, the user equipment 1 passes the detection. And analyzing the electrical signal generated by the user pressing the touch screen to obtain an input key sequence input by the user.
  • step S2 the user equipment 1 performs a query in the candidate lexicon according to the input key sequence and the keyboard layout of the keyboard to acquire one or more candidate characters corresponding to the input key sequence. string. Specifically, in step S2, the user equipment 1 combines the keyboard layout of the user equipment 1 according to the input key sequence acquired in step S1. A query is made in the candidate lexicon to obtain one or more candidate strings corresponding to the input key sequence.
  • the manner in which the user equipment 1 acquires the candidate character string includes, but is not limited to: 1) a character mapping table based on a preset keyboard layout, such as “g ⁇ f, h, t, v", “s ⁇ a, d , w, x”, etc., get one or more mapping strings corresponding to the input key sequence, such as obtaining the mapping strings “hongs”, “fongs”, “honha”, “tonfd”, etc.
  • the keys corresponding to the first extended character strings are closer to the input key sequence, such as obtaining the first extended character string “sun” according to the input key sequence “sin” , “san” and “son”, according to the keyboard layout, select candidate strings from these first extended strings, such as “u”, “0” and “ ⁇ adjacent to the button position, then select "sun” and “”son” as a candidate string, here, "distance” refers to the first extended word
  • the distance between the corresponding key in the input keyboard and the corresponding key position of the input key sequence in the input keyboard 3) acquiring one or more corresponding to the input key sequence according to the keyboard layout of the user input keyboard
  • the second extended string such as the "k” in the QWERTY full keyboard
  • the above three methods for acquiring candidate character strings are not mutually exclusive, but can be implemented together in the process of acquiring candidate character strings at a time.
  • the user equipment 1 obtains the corresponding candidate character string according to the above-mentioned three types of obtaining candidate character strings according to the input key sequence of the user, and combines the candidate character strings obtained by the three methods to obtain the final input with the user.
  • a candidate string corresponding to the key sequence is not mutually exclusive, but can be implemented together in the process of acquiring candidate character strings at a time.
  • the candidate vocabulary includes a local vocabulary and a network vocabulary; and, the candidate vocabulary further includes a Chinese vocabulary, English vocabulary, other language vocabulary, and a multi-language vocabulary.
  • the user equipment 1 is continuously operated in steps S1 and S2. Specifically, in step S1, the user equipment 1 is configured to acquire an input key sequence input by the user via the keyboard of the user equipment 1; in step S2, the user equipment 1 is configured to combine the keyboard of the keyboard according to the input key sequence. Layout, querying in the candidate lexicon to obtain one or more candidate character strings corresponding to the input key sequence.
  • “continuous" means that the user equipment 1 performs the acquisition of the input key sequence and the acquisition of the candidate character string according to the set or real-time adjusted working mode requirements in steps S1 and S2, respectively. Until the user stops inputting the input key sequence via the keyboard of the user device 1 for a long time.
  • FIG. 5 is a flow diagram of a method for obtaining a candidate character string corresponding to a sequence of input keys, in accordance with a preferred embodiment of the present invention.
  • the user equipment 1 performs a query in the candidate vocabulary according to the input key sequence to obtain one or more first extended character strings corresponding to the input key sequence;
  • the user equipment 1 selects the one or more candidate character strings from the one or more first extended character strings according to the keyboard layout.
  • step S21 the user equipment 1 performs a query in the candidate lexicon according to the input key sequence to acquire one or more first extended strings corresponding to the input key sequence. .
  • the user equipment 1 performs a query in the candidate lexicon according to the input key sequence input by the user via the keyboard of the user equipment 1 to acquire one or more corresponding to the input key sequence.
  • the first extended string For example, assuming that the key sequence "sin" is input, the candidate lexicon is the local vocabulary of the user device 1, and the user device 1 performs an extended query on the "sm" in the candidate lexicon, and obtains the corresponding first extended string "sun”. , "san” and "son”.
  • first extended character string is only an example, and other existing or future possible ways of obtaining the first extended character string may be applicable to the present invention, and should also be included in the present invention. Within the scope of protection, and is hereby incorporated by reference.
  • step S22 the user equipment 1 according to the keyboard layout, from the one Or selecting the one or more candidate strings from the plurality of first extension strings. Specifically, in step S22, the user equipment 1 selects one or more corresponding to the keyboard layout from the one or more first extended character strings obtained in step S21 ' according to the keyboard layout of the user input keyboard. Candidate strings.
  • the user device 1 obtains the first extended character strings “sun”, “san”, and “son” in step S21, and the user's input keyboard is a QWERTY full keyboard, "u", "0", and ""
  • the i" button positions are adjacent, and in step S22, the user device 1 selects "sun” and "son” as the candidate character strings from these first extended character strings.
  • the manner of obtaining the candidate character string is only an example. Other existing or future possible methods for obtaining the candidate character string may be applicable to the present invention and should also be included in the scope of the present invention. And is included here by reference.
  • step S23 the user equipment 1 acquires one or more second extended character strings corresponding to the input key sequence according to the input key sequence, in combination with the keyboard layout; in step S24" The user equipment 1 performs a matching query on the one or more second extended character strings in the candidate vocabulary to obtain the one or more candidate character strings.
  • step S23 the user equipment 1 acquires one or more second extended character strings corresponding to the input key sequence according to the input key sequence, in combination with the keyboard layout.
  • step In S23 the user equipment 1 acquires one or more second extended character strings corresponding to the input key sequence in accordance with the input key sequence input by the user via the user equipment 1, in conjunction with the keyboard layout of the user equipment 1.
  • the keyboard of the user device 1 is a QWERTY full keyboard, according to the keyboard layout, "u” is adjacent to the "y", “ ⁇ , "J” button positions, according to which, the user device 1 Obtain the second extended string “ying”, “iing", “jing”.
  • the user device 1 is a mobile phone
  • the keyboard is a normal mobile phone keyboard (keypad), that is, one button corresponds to more
  • the button "ghi" and the button “mno” are adjacent to the button “jkl” button position, whereby the user device 1 obtains the second extended character string "gent", “hent”, “"ient”,”ment”,”nent”,”oent”,”kent”,”lent” O
  • the manner of obtaining the second extended string is only an example. Other existing or future possible methods for obtaining the second extended string may be applicable to the present invention, and should also be included in the scope of protection of the present invention. Included here.
  • step S24 the user equipment 1 performs a matching query on the one or more second extended character strings in the candidate vocabulary to obtain the one or more candidate character strings.
  • step S24 the user equipment 1 performs a matching query on the one or more second extended character strings obtained in step S23" in the candidate lexicon to acquire one or more candidate character strings. For example, connect For example, the user equipment 1 obtains the second extended character string "ying”, “iing", “jing” in step S23", and in step S24", the user equipment 1 performs these second extended character strings in the candidate lexicon. Matching the query to obtain the candidate strings "ymg" and "jing".
  • the user device 1 obtains the second extended string "gent”, “hent”, “ient”, “ment”, “nent” in step S23” ", “oent”, “kent”, “lent”, in step S24", the user equipment 1 performs a matching query on the second extended character strings in the candidate lexicon, and obtains candidate strings "gent", “hent”. , “kent”, “lent”.
  • the manner of obtaining the candidate character string is only an example, and other existing or future possible methods for obtaining the candidate character string are applicable to the present invention, and are also included in the scope of the present invention and are incorporated herein by reference. .
  • the process further includes step S3 (not shown).
  • step S3 the user equipment 1 filters the one or more candidate strings according to the query assistance information. To get one or more candidate strings after filtering.
  • the user equipment 1 is configured to acquire an input key sequence input by the user via the keyboard of the user equipment 1; in step S2, the user equipment 1 is used to The input key sequence is combined with the keyboard layout of the keyboard to perform a query in the candidate lexicon to obtain one or more candidate character strings corresponding to the input key sequence; the specific process is as described above with reference to FIG.
  • the user equipment 1 in the described embodiment is identical to the process performed in step S2 in step S1, and is included herein by way of citation for the sake of brevity.
  • step S3 the user equipment 1 queries the auxiliary information, such as the user.
  • the candidate character strings obtained by the user equipment 1 according to the input key sequence "sm" are "son” and "sun", assuming that the current application of the user equipment 1 is a contact input of the mobile phone book, and Chinese input
  • step S3 the user equipment 1 filters the candidate character strings "son” and “sun” according to the current application, and obtains the filtered candidate character string "sun”.
  • the input key sequence is "sin”
  • the user's input keyboard is a QWERTY full keyboard
  • the user device 1 records "son” according to the user's previous input history
  • the input key sequence "sin” and the keyboard layout of the input keyboard If the adjacent keys of "i" are "u” and "0", query in the candidate lexicon to get the candidate string "son”.
  • the manner of filtering the candidate character string according to the query auxiliary information is only an example, and other existing or future possible methods for screening candidate character strings according to the query auxiliary information may be applicable to the present invention, and It is intended to be included within the scope of the invention and is hereby incorporated by reference.
  • the query assistance information includes, but is not limited to, at least one of the following:
  • the current application of the user equipment 1 refers to an application currently running by the user equipment 1.
  • the user equipment 1 filters one or more candidate character strings acquired in step S2 according to the currently running application, or according to the currently running application, and the input key sequence input by the user. And the keyboard layout of the input keyboard, performing a matching query in the candidate lexicon, and filtering to obtain the candidate character string, so that the candidate character string(s) not only satisfy the expansion of the input key sequence based on the keyboard layout, but also conform to the user equipment 1 Current application.
  • the input key sequence input by the user is "sin”
  • the current application of the user device 1 is the contact input of the phone book, and the user device 1 obtains the corresponding in step S2.
  • the candidate strings are "son” and "sun”.
  • the user equipment 1 filters the candidate strings according to the current application to obtain the candidate character string "sun”.
  • the user characteristic information of the user includes the user's input history record, the user's input preference setting, and the like.
  • the user device 1 is based on the user's user feature information, such as the user's input history record, the user's input preference setting, and the like. Filtering one or more candidate character strings obtained in step S2, or combining with step S2, according to the user characteristic information of the user, and the input key sequence input by the user and the keyboard layout of the input keyboard, in the candidate lexicon A matching query is performed, and the candidate string is obtained by filtering, so that the candidate character string(s) not only satisfy the extension of the input key sequence based on the keyboard layout, but also conform to the user characteristic information of the user.
  • the user's input preference is set to "m ⁇ n", that is, the user habitually hits “m” when tapping "n”, assuming the input key sequence is “ming”, and the input keyboard is arranged in alphabetical order.
  • the full keyboard, the user device 1 according to the input preference setting, and the input key sequence and the keyboard layout, such as "m” adjacent keys are “n” and “ ⁇ , the string that defines the extension of "ming” is “ning”, and query the "ning" in the candidate lexicon to obtain the candidate string "nmg".
  • the prior input information of the input key sequence includes the character string and the entry entered before the input key sequence recorded in the buffer of the user equipment 1, and in step S3, the user equipment 1 inputs the information according to the (some) prior input
  • the contextual semantic analysis between the candidate strings obtained in step S2, the candidate string(s) are filtered, or combined with step S2, based on the prior input information, and the input entered by the user.
  • the key sequence and the keyboard layout of the input keyboard are matched query and context semantic analysis in the candidate lexicon, and the candidate strings are selected by filtering, so that the candidate string(s) not only satisfy the expansion of the input key sequence based on the keyboard layout, It also conforms to the semantics of prior input information.
  • the input key sequence input by the user is "ahiwuduo"
  • the user equipment 1 obtains corresponding candidate character strings in step S2, such as “shiwusuo”, “zhisuduo”, “shiwuduo”, “zhiwusuo”, in the cache of the user device 1.
  • Recording the entry that the user inputs before inputting the input key sequence is "lawyer”, in step S3, the user equipment 1 performs context semantic analysis on the prior input information and the candidate character string respectively, and determines that the candidate character string is "Shiwusuo".
  • the semantic analysis result corresponding to the candidate character string includes a result obtained by semantic analysis of the candidate character string itself, such as whether the candidate character string is meaningful, and in step S3, the user equipment 1 acquires the same in step S2.
  • the candidate string is semantically analyzed, and the candidate string is filtered according to the semantic analysis result(s), or combined with step S2, and matched in the candidate lexicon according to the input key sequence input by the user and the keyboard layout of the input keyboard.
  • the query performs semantic analysis on the candidate string, and according to the result of the semantic analysis, the candidate string is obtained by screening, so that the candidate string(s) not only satisfy the extension of the input key sequence based on the keyboard layout, but also the candidate string itself.
  • the input key sequence input by the user is "sime people"
  • the input keyboard of the user is a QWERTY full keyboard
  • the user device 1 is based on the input key sequence and the keyboard layout in step S2, such as the adjacent key of "i”"o”,”u”, etc., the adjacent keys of "s” are “d”, “w”, "x”, etc., and are queried in the candidate lexicon to obtain candidate strings “some people”, “dime people””and” dome people”, in step S3, the user equipment 1 separately performs semantic analysis on the candidate character strings, and determines candidate character strings according to the corresponding semantic analysis results, such as "some people”
  • the distance threshold of the candidate character string and the input key sequence includes a preset maximum value of the number of characters different between the candidate character string and the input key sequence, and the corresponding key position and input button of the character in the candidate character string in the input keyboard The preset maximum distance between the corresponding key positions in the input keyboard.
  • the user equipment 1 filters the candidate character string acquired in step S2 according to the distance threshold, or combines with step S2, according to the distance threshold, and the input key sequence input by the user and the input keyboard.
  • the user input key sequence is "ahurufa", assuming that the preset distance threshold is 1, where the distance threshold refers to a preset maximum value of the number of characters between the candidate character string and the input key sequence, that is, the candidate string is at most only It can be different from the input key sequence by one character.
  • the user's input keyboard is a QWERTY full keyboard.
  • the user equipment 1 follows the distance threshold, and the input key sequence input by the user and the keyboard layout of the input keyboard, such as "a" adjacent.
  • the keys are “z", “s”, etc., the adjacent keys of "u” are “i”, “y” Etc., query in the candidate lexicon, and filter the candidate strings into “shurufs” and "zhurufa”.
  • the user input key sequence is "jind”
  • the distance threshold is 1, where the distance threshold refers to the corresponding button position of the character in the input keyboard and the corresponding button position of the input key sequence in the input keyboard.
  • the preset maximum distance between the input keyboard is QWERTY full keyboard, the user equipment 1 according to the distance threshold, and the input key sequence input by the user and the keyboard layout of the input keyboard, determining that the adjacent key of "j" is “k”, “m”, “h”, etc., "The adjacent keys of ⁇ are “u”, etc., and are queried in the candidate lexicon to obtain candidate strings “kind”, “mind”, “hind”, “jund”.
  • the above-mentioned query auxiliary information can be used not only for screening candidate strings, but also for combining candidate pairs or even combining them for screening candidate strings.
  • the above-mentioned query auxiliary information is only an example, and other existing or future query auxiliary information may be applicable to the present invention, and should also be included in the scope of the present invention, and The way is included here.
  • the user characteristic information includes, but is not limited to, at least one of the following:
  • the input history of the user including but not limited to the input key sequence of the user history input, the user's selection operation of the candidate character string, the mapping relationship between the input key sequence of the user history input, and the candidate string selected by the user, etc. If the input key sequence is "sm” and the corresponding candidate strings are “sun” and “son”, the user selects the candidate character string “sun”, and the user device 1 establishes a mapping relationship between "sin” and "sun”.
  • the user's input preference settings including but not limited to the mapping relationship between the characters specially set by the user according to their input habits, such as "m ⁇ n", and the user's preference for the input method, such as the high priority of the English input. The priority entered in Chinese.
  • the process further comprises a step S4 (not shown), in which the user equipment 1 provides the one or more candidate character strings to the user equipment 1.
  • step S1 the user equipment 1 is configured to acquire an input key sequence input by the user via the keyboard of the user equipment 1; in step S2, the user equipment 1 uses Querying in the candidate lexicon according to the input key sequence, in combination with the keyboard layout of the keyboard, to obtain one or more candidate character strings corresponding to the input key sequence; the specific process and the foregoing reference figure
  • the user equipment 1 in the described embodiment is the same as the one performed in step S2 in step S1, and is included herein for the sake of brevity and is not to be construed as a reference.
  • step S4 the user equipment 1 randomly or in accordance with one or more candidate character strings acquired in step S2, such as the word frequency of the candidate character string from high to low, the order of ranking in the candidate lexicon, and the like. , provided to the user device 1.
  • the candidate character strings obtained by the user equipment 1 are "son” and “sun”, and the word frequency in the candidate lexicon according to the candidate character strings, such as the word frequency of "sun” is higher than the word frequency of "son”.
  • the user equipment 1 supplies the candidate character strings to the user equipment 1 in the order of word frequency from high to low, that is, "sun", "son”.
  • the user equipment 1 may also provide one or more candidate character strings that are filtered and obtained in step S3 to the user equipment 1 in the above manner.
  • the manner of providing the candidate character string is only an example, and other existing or future possible ways of providing the candidate character string may be applicable to the present invention and should also be included in the scope of the present invention. And is included here by reference.
  • step S4 the user equipment 1 provides a candidate character string.
  • the user can select the candidate character string that he/she wishes to input through the user equipment 1 for acquiring the corresponding candidate term.
  • This step can also implement an extended reminder of the input key sequence input by the user.
  • the candidate character string is an input error reminder for the input key sequence, which can enable the user to understand the wrong input habit in time.
  • the process further includes step S5 (not shown), and in step S5, the user equipment 1 updates the candidate term store according to the user's selection operation of the one or more candidate character strings.
  • the user equipment 1 according to the user is in step S4.
  • the user equipment 1 provides candidate character strings "son” and "sun”, and the user performs a selection operation on the candidate character strings via a keyboard or a mouse, such as the user selecting the candidate character string "son”, in step S5.
  • the user equipment 1 updates the word frequency of "son” in the candidate lexicon according to the selection operation of the user, that is, the word frequency of "son” is increased by one.
  • the input key sequence is "sin”
  • the corresponding candidate character strings provided by the user equipment 1 are “son” and “sun”
  • the user selects the candidate character string "son”.
  • the user equipment 1 According to the selection operation, the mapping relationship between the input key sequence and the candidate character string in the candidate lexicon is updated, that is, "sin” and "son”.
  • the manner of updating the candidate vocabulary is only an example, and other existing or future possible candidates for updating the vocabulary may be applicable to the present invention and should also be included in the scope of the present invention. And is included here by reference.
  • the process further includes steps S6 (not shown) and S7 (not shown), and in step S6, the user equipment 1 is based on the one or more candidate strings. Performing a matching query in the matching thesaurus to obtain one or more candidate terms corresponding to the one or more candidate character strings; in step S7, the user equipment 1 will use the one or more candidate words The strip is provided to the user device 1.
  • step S1 the user equipment 1 is configured to acquire an input key sequence input by the user via the keyboard of the user equipment 1; in step S2, the user equipment 1 uses Querying in the candidate lexicon according to the input key sequence, in combination with the keyboard layout of the keyboard, to obtain one or more candidate character strings corresponding to the input key sequence; the specific process and the foregoing reference figure
  • the user equipment 1 in the described embodiment is the same as the one performed in step S2 in step S1, and is included herein by way of citation for the sake of brevity.
  • step S6 the user equipment 1 performs a matching query in the matching thesaurus according to one or more candidate character strings acquired in step S2 to obtain one or more corresponding to the candidate character string(s).
  • Candidate terms For example, the user equipment 1 acquires the candidate character strings "sun", "son” in step S2, and the user equipment 1 pairs the candidate character strings in step S6.
  • Matching queries are performed in the matching thesaurus to obtain corresponding candidate terms "sun” and "son”.
  • the user equipment 1 may also perform a matching query on the candidate character string corresponding to the user's selection operation in the matching vocabulary to obtain the corresponding candidate term; Sun", the term acquisition means only performs a matching query on the candidate string "sun” in the matching thesaurus to obtain the corresponding candidate term "sun".
  • the matching thesaurus can be independent of the candidate thesaurus, or can be combined into one thesaurus.
  • step S7 the user equipment 1 provides one or more candidate terms it acquired in step S6 to the user equipment 1.
  • the user equipment 1 acquires the candidate terms “sun” and “son” in step S6, and the frequency of use of the user in the candidate lexicon according to the candidate terms, such as the "sun” user frequency is higher than "son” The user uses the frequency.
  • the user equipment 1 provides the candidate terms to the user equipment 1 in descending order of user frequency, that is, "sun” and "son”.
  • the user equipment 1 provides one or more candidate terms that it acquires, on the one hand, for the user to select a candidate term that he or she wishes to input; on the other hand, an extension of the candidate term that the user can select is also implemented, Further, the error correction process when the user inputs an error is reduced, and the accuracy of the user input is also made possible.
  • the process further includes a step S8 (not shown), in step S8, the user equipment 1 performs a matching query in the matching vocabulary according to the input key sequence to obtain a sequence with the input key Corresponding one or more matching terms; in step S7, the user equipment 1 provides the one or more matching terms and the one or more candidate terms to the user equipment 1.
  • step S8 the user equipment 1 performs a matching query in the matching vocabulary according to the input key sequence acquired in step S1 to obtain one or more matching terms corresponding to the input key sequence.
  • the user equipment 1 also supplies one or more matching terms it acquired in step S8 and its candidate terms acquired in step S6 to the user equipment 1, such as matching entries.
  • Merged with candidate terms enter one Steps, it is also possible to combine a plurality of matching terms with the highest word frequency and a plurality of candidate words with the highest word frequency, and provide the user equipment 1 according to the word frequency of the combined words, the frequency of use by the user, and the like.
  • the candidate lexicon and the matching vocabulary are a thesaurus
  • the user equipment 1 performs a matching query on the input key sequence in the lexicon, and obtains a corresponding matching term "thinking", “four years”, “raging”, “private thoughts”, and obtaining corresponding candidate strings, such as “son” and “sun”, thereby obtaining corresponding candidate terms "son", "sun”, user device 1 according to
  • the word frequency of the matching term and the candidate term in the lexicon, and the matching vocabulary and the candidate vocabulary are provided to the user equipment 1 in descending order of word frequency, such as "thinking", "sun", “four” Years, “raging”, "son”, “private thoughts”.
  • the user equipment 1 provides the matching term and the candidate term, on the one hand, providing the user with more choices; on the other hand, in the case that the matching term is not the term that the user wishes to input, based on the input button
  • the extension of the sequence provides candidate terms, which reduces the number of times users need to re-enter, and improves user input efficiency and input accuracy.
  • the process further includes a step S9 (not shown), in step S9, the user equipment 1 acquires a selection operation of the one or more candidate character strings by the user; in step S6, the user equipment 1 further a candidate character string corresponding to the selection operation, performing a matching query in the matching vocabulary to obtain one or more candidate terms corresponding to the candidate character string corresponding to the selection operation.
  • step S9 the user equipment 1 acquires a selection operation of the user for one or more candidate character strings; in step S6, the user equipment 1 further selects the matching word according to the candidate character string corresponding to the selection operation.
  • a matching query is performed in the library to obtain one or more candidate terms corresponding to the candidate string corresponding to the selection operation.
  • the user device 1 acquires the user's selection operation for these candidate character strings, such as the user selecting the candidate character string "sun” through the keyboard; the user device 1 is based on “sun” Matching queries are performed in the matching thesaurus to obtain candidate terms "sun", “loss”, “bamboo” and the like.
  • Matching queries are performed in the matching thesaurus to obtain candidate terms "sun”, “loss”, “bamboo” and the like.
  • the process further includes a step S10 (not shown), in which the user device 1 updates the matching thesaurus according to the user's selection operation of the one or more candidate terms.
  • the user equipment 1 updates the matching vocabulary according to the user's selection operation of the candidate term, such as updating the word frequency of the candidate term selected by the user in the matching vocabulary, updating the input key sequence and The mapping relationship of candidate terms selected by the user, and the like.
  • the user device 1 obtains the corresponding candidate strings “son” and “sun”, and the corresponding candidate terms “sun”, “son”, and provides the two candidate terms to User equipment 1 , the user selects a candidate term “sun” via the user device 1 , and the user device 1 updates the word frequency of the “sun” in the matching vocabulary according to the selection operation, and adds 1 to the word “sun”, and updates the input at the same time.

Description

获取与输入按键序列相对应的候选字符串的方法与设备 技术领域
本发明涉及输入法技术领域， 尤其涉及一种获取与输入按键序列 相对应的候选字符串的技术。 背景技术
随着输入法技术的不断发展， 用户对使用输入法进行高效准确输 入的需求也越来越强烈。 现有的输入法， 通常基于用户输入的输入按 键序列， 在词库中进行查询， 获得相应的匹配词条。 进一步地， 现有 的输入法结合用户的语言习惯等建立模糊音规则， 如将" Z"映射为
"zh" , 以对用户输入的输入按键序列进行扩展， 如用户输入" zyao" , 可得到匹配词条"紫瑶"、 "则要 "等， 根据该模糊音规则扩展后， 还可 获得相应的候选字符串 "zhyao" , 进而获得候选词条， 如"重要"、 "摘 要"、 "主要 "等。
然而， 现有的输入法未能考虑由于用户敲击键盘的习惯等因素， 造成的输入错误， 以及基于键盘布局的输入按键序列的扩展。 用户在 进行输入时常常错按其所需按键的相邻按键， 从而用户设备难以获得 用户希望输入的字符串， 使得基于该字符串查询词库以获得候选词条 的过程更无从谈起。 相应地， 这也使得用户的输入速度和输入准确性 都受到了极大的限制， 同时降低了用户输入效率和用户使用体验。
因此， 如何提供一种获取与输入按键序列相对应的候选字符串的 方法或设备， 使得用户设备能够更快速、 更准确的获取用户希望输入 的字符串， 成为目前亟待解决的问题之一。 发明内容
本发明的目的是提供一种获取与输入按键序列相对应的候选字符 串的方法与设备。 根据本发明的一个方面， 提供了一种获取与输入按键序列相对应 的候选字符串的方法， 其中， 该方法包括以下步骤：
a获取用户经由用户设备的键盘输入的输入按键序列；
b根据所述输入按键序列， 结合所述键盘的键盘布局， 在候选词库 中进行查询， 以获取与所述输入按键序列相对应的一个或多个候选字符 串。
根据本发明的另一方面， 还提供了一种获取与输入按键序列相对 应的候选字符串的设备， 其中， 该设备包括：
输入获取装置， 用于获取用户经由用户设备的键盘输入的输入按 键序列；
字符串获取装置， 用于根据所述输入按键序列， 结合所述键盘的 键盘布局， 在候选词库中进行查询， 以获取与所述输入按键序列相对 应的一个或多个候选字符串。
与现有技术相比， 本发明根据用户输入的输入按键序列， 以及用 户的输入键盘的键盘布局， 例如 QWERTY 全键盘、 普通手机键盘 ( keypad ) 、 按字母顺序排列的全键盘， 以及任何自定义布局的键盘 等的键盘布局， 在候选词库中进行查询， 以获取相应的候选字符串， 从而实现对用户输入的输入按键序列进行基于按键在键盘上分布方 式的扩展， 并获得相应的用于获取候选词条的候选字符串。 本发明减 少了由于用户敲击键盘的习惯等因素所造成的输入错误， 并且加入了 基于键盘布局的输入按键序列的扩展， 从而有效地提高用户的输入效 率并提升了用户的使用体验。 附图说明
通过阅读参照以下附图所作的对非限制性实施例所作的详细描述， 本发明的其它特征、 目的和优点将会变得更明显：
图 1示出根据本发明一个方面的用于获取与输入按键序列相对应 的候选字符串的设备示意图； 图 2示出根据本发明一个优选实施例的用于获取与输入按键序列 相对应的候选字符串的设备示意图；
图 3示出根据本发明另一优选实施例的用于获取与输入按键序列 相对应的候选字符串的设备示意图；
图 4示出根据本发明另一个方面的用于获取与输入按键序列相对 应的候选字符串的方法流程图；
图 5根据本发明一个优选实施例的用于获取与输入按键序列相对 应的候选字符串的方法流程图；
图 6根据本发明另一优选实施例的用于获取与输入按键序列相对 应的候选字符串的方法流程图。
附图中相同或相似的附图标记代表相同或相似的部件。 具体实施方式
下面结合附图对本发明作进一步详细描述。
图 1为根据本发明一个方面的设备示意图， 示出获取与输入按键 序列相对应的候选字符串的用户设备。 其中， 用户设备 1 包括输入获 取装置 11、 字符串获取装置 12。 具体地， 输入获取装置 11获取用户经 由用户设备 1 的键盘输入的输入按键序列； 字符串获取装置 12根据所 述输入按键序列， 结合所述键盘的键盘布局， 在候选词库中进行查询， 以获取与所述输入按键序列相对应的一个或多个候选字符串。 在此， 用 户设备 1可以是任何一种能与用户通过键盘进行人机交互的电子产品， 例如计算机、 手机、 PDA、 平板电脑、 PSP、 掌上电脑 PPC等； 其中， 键盘可以是任何一种物理键盘或虚拟键盘， 包括但不限于 QWERTY全 键盘、 普通手机键盘（keypad ) 、 按字母顺序排列的全键盘， 以及任何 自定义布局的虚拟键盘。 本领域技术人员应能理解上述用户设备以及键 盘仅为举例， 其他现有的或今后可能出现的用户设备或者键盘如可适用 于本发明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
更具体地， 输入获取装置 11获取用户经由用户设备 1 的键盘输入 的输入按键序列。 具体地， 输入获取装置 11 通过检测并解析用户按压 用户设备 1的触摸屏上的虚拟键盘或者敲击物理键盘等产生的电信号， 获取该用户经由该用户设备 1的键盘输入的输入按键序列。 例如， 假设 用户设备 1为带有触摸屏的电子产品， 相应的虚拟键盘为 QWERTY全 键盘， 用户通过按压该 QWERTY全键盘中的按键， 向该用户设备 1进 行输入， 相应地， 输入获取装置 11 通过检测并解析该用户按压该触摸 屏所产生的电信号， 获取该用户输入的输入按键序列。 本领域技术人员 应能理解上述获取输入按键序列的方式仅为举例， 其他现有的或今后可 能出现的获取输入按键序列的方式如可适用于本发明， 也应包含在本发 明保护范围以内， 并以引用方式包含于此。
字符串获取装置 12根据所述输入按键序列，结合所述键盘的键盘布 局， 在候选词库中进行查询， 以获取与所述输入按键序列相对应的一个 或多个候选字符串。 具体地， 字符串获取装置 12根据输入获取装置 11 获取的输入按键序列， 结合用户设备 1的键盘布局， 在候选词库中进行 查询， 以获取与该输入按键序列相对应的一个或多个候选字符串。在此， 字符串获取装置 12获取候选字符串的方式包括但不限于： 1 )根据预设 的基于键盘布局的字符映射表， 如" g→f、 h、 t、 v", "s→a、 d、 w、 x" 等， 获取与输入按键序列相对应的一个或多个映射字符串， 如根据 "gongs"获得映射字符串" hongs"、 "fongs"、 "honha"、 "tonfd"等， 再在候 选词库中对这些映射字符进行匹配查询，以获得候选字符串，如" hongs"; 2 )根据用户输入的输入按键序列， 在候选词库中进行查询， 以获取与 输入按键序列相对应的一个或多个第一扩展字符串， 这些第一扩展字符 串所对应的按键与该输入按键序列的距离较近， 如根据输入按键序列 "sin"获得第一扩展字符串" sun"、 "san"和" son", 再根据键盘布局， 从这 些第一扩展字符串中选择候选字符串， 如 "u"、 "0"与" Γ按键位置相邻， 则选择 "sun"和" son"作为候选字符串，在此， "距离"指第一扩展字符串中 字符在输入键盘中对应的按键与输入按键序列在该输入键盘中对应的 按键位置之间的距离； 3 )根据用户输入键盘的键盘布局， 获取与输入 按键序列相对应的一个或多个第二扩展字符串， 如 QWERTY全键盘中 "k"与" i"、 "0"、 T、 "m"、 "j"按键位置相邻, "rkse"对应的第二扩展字 符串为 "rise"、 "rose", "rise", "rmse"、 "rjse", 再在候选词库中对这些第 二扩展字符串进行匹配查询， 获得候选字符串， 如" rise"和" rose"。
需要说明的是， 本领域技术人员应能理解， 上述 3种获取候选字符 串的方式之间不是互相排斥的， 而是可以共同在一次获取候选字符串的 过程中实现的。 例如， 字符串获取装置 12根据用户的输入按键序列， 分别按照上述 3种获取候选字符串的方式获取相应的候选字符串， 并将 3 种方式所获得的候选字符串合并， 以获得最终与用户的输入按键序列 相对应的候选字符串。
本领域技术人员还应能理解上述获取候选字符串的方式仅为举例， 其他现有的或今后可能出现的获取候选字符串的方式如可适用于本发 明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。优选地， 候选词库包括本地词库和网络词库； 并且， 候选词库还包括中文词库、 英文词库、 其他语言词库， 以及多种语言混合的词库。
优选地， 输入获取装置 11和字符串获取装置 12之间是持续不断地 工作。 具体地， 输入获取装置 11用于获取用户经由用户设备 1 的键盘 输入的输入按键序列； 字符串获取装置 12用于根据所述输入按键序列， 结合所述键盘布局， 在候选词库中进行查询， 以获取与所述输入按键序 列相对应的一个或多个候选字符串。在此，本领域技术人员应理解"持续" 是指输入获取装置 11和字符串获取装置 12分别按照设定的或实时调整 的工作模式要求进行输入按键序列的获取、 以及候选字符串的获取， 直 至用户在较长时间内停止经由用户设备 1 的键盘进行输入按键序列的 输入。
图 2为根据本发明一个优选实施例的设备示意图， 示出获取与输 入按键序列相对应的候选字符串的用户设备。 其中， 字符串获取装置 12，包括第一扩展单元 121，和选择单元 122，。具体地，第一扩展单元 121， 根据所述输入按键序列， 在所述候选词库中进行查询， 以获取与所述输 入按键序列相对应的一个或多个第一扩展字符串； 选择单元 122，根据所 述键盘布局， 从所述一个或多个第一扩展字符串中选择所述一个或多个 候选字符串。
更具体地， 第一扩展单元 121，根据所述输入按键序列， 在所述候选 词库中进行查询， 以获取与所述输入按键序列相对应的一个或多个第一 扩展字符串。 具体地， 第一扩展单元 121，根据用户经由用户设备 1的键 盘输入的输入按键序列， 在候选词库中进行查询， 以获取与该输入按键 序列相对应的一个或多个第一扩展字符串。 例如， 假设输入按键序列 "sin", 候选词库为用户设备 1的本地词库， 第一扩展单元 121，在该候选 词库中对" sin"进行扩展查询， 得到相应的第一扩展字符串 "sun"、 "san" 和" son"。本领域技术人员应能理解上述获取第一扩展字符串的方式仅为 举例， 其他现有的或今后可能出现的获取第一扩展字符串的方式如可适 用于本发明，也应包含在本发明保护范围以内，并以引用方式包含于此。
随后， 选择单元 122，根据所述键盘布局， 从所述一个或多个第一扩 展字符串中选择所述一个或多个候选字符串。 具体地， 选择单元 122，根 据用户输入键盘的键盘布局， 从第一扩展单元 121，获取的一个或多个第 一扩展字符串中选择与该键盘布局相对应的一个或多个候选字符串。 例 如， 接上例， 第一扩展单元 121，获得第一扩展字符串 "sun"、 "san" 和 "son", 用户的输入键盘为 QWERTY全键盘， "u"、 "o"与" i"按键位置相 邻， 选择单元 122，从这些第一扩展字符串中选择" sun"和" son"作为候选 字符串。 本领域技术人员应能理解上述获取候选字符串的方式仅为举 例， 其他现有的或今后可能出现的获取候选字符串的方式如可适用于本 发明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
图 3为根据本发明另一优选实施例的设备示意图， 示出获取与输 入按键序列相对应的候选字符串的用户设备。 其中， 字符串获取装置 12"包括第二扩展单元 123"和查询单元 124"。具体地，第二扩展单元 123" 根据所述输入按键序列， 结合所述键盘布局， 获取与所述输入按键序列 相对应的一个或多个第二扩展字符串； 查询单元 124"在所述候选词库中 对所述一个或多个第二扩展字符串进行匹配查询， 以获取所述一个或多 个候选字符串。
更具体地， 字符串获取装置 12"的第二扩展单元 123"根据所述输 入按键序列， 结合所述键盘布局， 获取与该输入按键序列相对应的一个 或多个第二扩展字符串。 具体地， 第二扩展单元 123"根据用户经由用 户设备 1输入的输入按键序列， 结合该用户设备 1的键盘布局， 获取与 该输入按键序列相对应的一个或多个第二扩展字符串。 例如， 假设输入 按键序列" uing",用户设备 1的键盘为 QWERTY全键盘，根据该键盘布 局， "u"与" y"、 "i"、 "j"按键位置相邻, 据此， 第二扩展单元 123"获得 第二扩展字符串" ying"、 "iing"、 "jing"。 又如， 假设输入按键序列" jent", 用户设备 1为手机， 其键盘为普通手机键盘， 即一个按键对应多个字母 的键盘， 根据该键盘布局， 按键" ghi"、 按键" mno"与按键 "jkl"按键位置 相邻， 据此， 第二扩展单元 123"获得第二扩展字符串 "gent"、 "hent"、 "ient"、 "ment"、 "nent"、 "oent"、 "kent"、 "lent"。 本领域技术人员应能理 解上述获取第二扩展字符串的方式仅为举例， 其他现有的或今后可能出 现的获取第二扩展字符串的方式如可适用于本发明， 也应包含在本发明 保护范围以内， 并以引用方式包含于此。
随后，查询单元 124"在所述候选词库中对所述一个或多个第二扩展 字符串进行匹配查询， 以获取所述一个或多个候选字符串。 具体地， 查 询单元 124"在该候选词库中对从第二扩展单元 123"中获得的一个或多 个第二扩展字符串进行匹配查询， 以获取一个或多个候选字符串。例如， 接上例， 第二扩展单元 123"获得第二扩展字符串 "ying"、 "iing"、 "jing", 查询单元 124"在候选词库中对这些第二扩展字符串进行匹配查询，获得 候选字符串" ymg"和 "jmg"。 又如， 第二扩展单元 123"获得第二扩展字符 串" gent"、 "hent"、 "ient"、 "ment"、 "nent"、 "oent"、 "kent"、 "lent", 查 询单元 124"在候选词库中对这些第二扩展字符串进行匹配查询，得到候 选字符串" gent"、 "hent"、 "kent"、 "lent"。 本领域技术人员应能理解上述 获取候选字符串的方式仅为举例， 其他现有的或今后可能出现的获取候 选字符串的方式如可适用于本发明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
在一个优选实施例中（参照图 1 ) , 用户设备 1还包括筛选装置（未 示出） ， 该筛选装置才艮据查询辅助信息， 对所述一个或多个候选字符串 进行筛选， 以获取筛选后的一个或多个候选字符串。 以下参照图 1对该 优选实施例进行详细描述， 其中， 输入获取装置 11 获取的用户经由用 户设备输入的输入按键序列， 字符串获取装置 12根据所述输入按键序 列来获取一个或多个候选字符串， 其具体过程与前述参照图 1所描述的 实施例中输入获取装置 11和字符串获取装置 12所执行的过程相同， 为 简明起见， 以引用方式包含于此， 而不做赞述。 具体地， 该筛选装置根 据查询辅助信息， 诸如该用户设备 1的当前应用、 该用户的用户特征信 息等， 对字符串获取装置 12 获取的一个或多个候选字符串进行筛选， 或者与字符串获取装置 12相结合获取及筛选候选字符串， 以获得筛选 后的一个或多个候选字符串。 例如， 字符串获取装置 12根据输入按键 序列" sin"获得的候选字符串为 "son"和" sun", 该用户设备 1 的当前应用 为手机电话薄的联系人输入， 且为中文输入， 根据该当前应用对候选字 符串" son"和" sun"进行筛选， 得到筛选后的候选字符串" sun"。 又如， 输 入按键序列为 "sin", 用户的输入键盘为 QWERTY全键盘， 筛选装置与 字符串获取装置 12相结合， 根据该用户之前的输入历史记录" son", 以 及该输入按键序列 "sin"和该输入键盘的键盘布局， 如" i"的相邻按键为 "u"和" 0" , 在候选词库中进行查询， 得到候选字符串" son"。 本领域技术 人员应能理解上述根据查询辅助信息筛选候选字符串的方式仅为举例， 其他现有的或今后可能出现的根据查询辅助信息筛选候选字符串的方 式如可适用于本发明， 也应包含在本发明保护范围以内， 并以引用方式 包含于此。
优选地， 查询辅助信息包括但不限于， 以下至少任一项：
1) 用户设备 1的当前应用；
2) 该用户的用户特征信息； 3) 所述输入按键序列的在先输入信息；
4) 所述候选字符串所对应的语义分析结果；
5) 所述候选字符串与所述输入按键序列的距离阈值。
具体地， 1 )用户设备 1的当前应用指用户设备 1当前运行的应用程 序。 筛选装置根据用户设备 1当前运行的应用程序， 对字符串获取装置 12获取的一个或多个候选字符串进行筛选，或者筛选装置与字符串获取 装置 12结合， 根据用户设备 1 当前运行的应用程序， 以及用户输入的 输入按键序列和输入键盘的键盘布局， 在候选词库中进行匹配查询， 筛 选获得候选字符串， 以使这（些）候选字符串不仅满足基于键盘布局的 对输入按键序列的扩展， 而且符合该用户设备 1的当前应用。 例如， 用 户输入的输入按键序列为" sin", 用户设备 1的当前应用为电话簿的联系 人输入， 字符串获取装置 12获取相应的候选字符串为" son"、 "sun", 筛 选装置根据该当前应用对这些候选字符串进行歸选， 获得候选字符串 "sun"。
2 )用户的用户特征信息包括用户的输入历史记录、用户的输入偏好 设置等， 筛选装置根据用户的用户特征信息，诸如用户的输入历史记录、 用户的输入偏好设置等， 对字符串获取装置 12 获取的一个或多个候选 字符串进行筛选， 或者筛选装置与字符串获取装置 12 结合， 根据用户 的用户特征信息， 以及用户输入的输入按键序列和输入键盘的键盘布 局， 在候选词库中进行匹配查询， 筛选获得候选字符串， 以使这（些） 候选字符串不仅满足基于键盘布局的对输入按键序列的扩展， 而且符合 该用户的用户特征信息。 例如， 用户的输入偏好设置为" m→n", 即该用 户习惯性的在敲击" n"时误敲击" m", 假设输入按键序列为 "ming", 输入 键盘为按字母顺序排列的全键盘， 筛选装置与字符串获取装置 12相结 合， 根据该输入偏好设置， 以及该输入按键序列和该键盘布局， 如" m" 的相邻按键为 "n"和" Γ, 确定对 "ming"进行扩展的字符串为 "ning", 并对 该" ning"在候选词库中进行查询 , 获得候选字符串 "ning"。
3 )输入按键序列的在先输入信息包括用户设备 1的緩存中记录的在 该输入按键序列之前输入的字符串和词条， 筛选装置根据这（些）在先 输入信息与字符串获取装置 12 获取的候选字符串之间的上下文语义分 析， 对这（些）候选字符串进行筛选， 或者筛选装置与字符串获取装置 12结合， 根据这（些）在先输入信息， 以及用户输入的输入按键序列和 输入键盘的键盘布局， 在候选词库中进行匹配查询和上下文语义分析， 筛选获得候选字符串， 以使这（些）候选字符串不仅满足基于键盘布局 的对输入按键序列的扩展， 而且符合在先输入信息的语义。 例如， 用户 输入的输入按键序列为" ahiwuduo",字符串获取装置 12获取相应的候选 字符串， 如" shiwusuo"、 "zhisuduo"、 "shiwuduo", "zhiwusuo", 用户设 备 1 的緩存中记录该用户在输入该输入按键序列之前输入的词条为 "律 师"，筛选装置对该在先输入信息分别与上述候选字符串进行上下文语义 分析， 确定候选字符串为" shiwusuo"。
4 )候选字符串所对应的语义分析结果包括通过对候选字符串本身的 语义分析所获得的结果， 如候选字符串是否有意义， 筛选装置对字符串 获取装置 12获取的候选字符串做语义分析， 根据这（些）语义分析结 果对候选字符串进行筛选， 或者筛选装置与字符串获取装置 12 结合， 根据用户输入的输入按键序列和输入键盘的键盘布局， 在候选词库中进 行匹配查询， 同时对候选字符串做语义分析， 根据该语义分析结果， 筛 选获得候选字符串， 以使这（些）候选字符串不仅满足基于键盘布局的 对输入按键序列的扩展， 而且候选字符串本身有意义。 例如， 假设用户 输入的输入按键序列为" sime people",该用户的输入键盘为 QWERTY全 键盘， 字符串获取装置 12基于该输入按键序列和键盘布局， 如" Γ的相 邻按键为" o"、 "u"等," s"的相邻按键为 "d"、 "w"、 "x"等， 在候选词库中 进行查询， 获得候选字符串" some people" , "dime people"和" dome people", 筛选装置分别对这些候选字符串做语义分析， 并根据相应的语 义分析结果确定候选字符串， 如" some people"
5 )候选字符串与该输入按键序列的距离阈值包括候选字符串与输入 按键序列之间不相同字符数的预设最大值， 以及候选字符串中字符在输 入键盘中对应的按键位置与输入按键序列在该输入键盘中对应的按键 位置之间的预设最大距离。 筛选装置按照该距离阈值， 对字符串获取装 置 12获取的候选字符串进行筛选， 或者筛选装置与字符串获取装置 12 结合， 根据该距离阈值， 以及用户输入的输入按键序列和输入键盘的键 盘布局，在候选词库中进行匹配查询，筛选获得候选字符串，以使这（些） 候选字符串不仅满足基于键盘布局的对输入按键序列的扩展， 而且不超 过候选字符串与该输入按键序列的距离阈值。 例如， 用户输入按键序列 为" ahurufa", 假设预设距离阈值为 1 , 在此， 距离阈值指候选字符串与 输入按键序列之间不相同字符数的预设最大值， 即候选字符串最多只能 与该输入按键序列有 1个字符不同， 用户的输入键盘为 QWERTY全键 盘， 筛选装置与字符串获取装置 12相结合， 按照该距离阈值， 以及用 户输入的输入按键序列和输入键盘的键盘布局， 如" a"的相邻按键为 "z"、 "s"等," u"的相邻按键为 "Γ、 "y"等， 在候选词库中进行查询， 筛选得到候 选字符串为" shurufs"、 "zhurufa"。 又如， 用户输入按键序列为 "jind", 假 设距离阈值为 1 , 在此， 距离阈值指候选字符串中字符在输入键盘中对 应的按键位置与输入按键序列在该输入键盘中对应的按键位置之间的 预设最大距离， 输入键盘为 QWERTY全键盘， 筛选装置与字符串获取 装置 12相结合， 按照该距离阈值， 以及用户输入的输入按键序列和输 入键盘的键盘布局， 确定" j"的相邻按键为 "k"、 "m"、 "h"等," i"的相邻按 键为" u"等，并在候选词库中进行查询，得到候选字符串 "kind"、 "mind", "hind", "jund"。
本领域技术人员应能理解， 上述各项查询辅助信息不仅可以单独用 于筛选候选字符串， 还可以两两结合、 甚至相互结合用于筛选候选字符 串。 本领域技术人员还应能理解上述各项查询辅助信息仅为举例， 其他 现有的或今后可能出现的查询辅助信息如可适用于本发明， 也应包含在 本发明保护范围以内， 并以引用方式包含于此。
更优选地， 用户特征信息包括但不限于， 以下至少任一项：
1 )该用户的输入历史记录； 2 )该用户的输入偏好设置。
具体地， 1 )用户的输入历史记录， 包括但不限于用户历史输入的输 入按键序列， 用户对候选字符串的选择操作， 用户历史输入的输入按键 序列与用户选择的候选字符串的映射关系等， 如输入按键序列为 "sm" , 相应的候选字符串为 "sun"和" son", 用户选择候选字符串 "sun", 据此用 户设备 1建立" sin"与 "sun"的映射关系。 2 )用户的输入偏好设置， 包括 但不限于用户根据其输入习惯而特别设置的字符之间的映射关系， 如 "m→n", 以及用户对输入法的偏好， 如英文输入的优先级高于中文输入 的优先级。 本领域技术人员应能理解， 上述各项用户特征信息不仅可以 单独用于筛选候选字符串， 还可以两两结合、 甚至相互结合用于筛选候 选字符串。 本领域技术人员应能理解上述各项用户特征信息仅为举例， 其他现有的或今后可能出现的用户特征信息如可适用于本发明， 也应包 含在本发明保护范围以内， 并以引用方式包含于此。
在另一优选实施例中 （参照图 1 ) , 用户设备 1还包括字符串提供 装置 （未示出） ， 该字符串提供装置将所述一个或多个候选字符串提供 至用户设备 1。 以下参照图 1对该另一优选实施例进行详细描述， 其中， 输入获取装置 11 获取的用户经由用户设备输入的输入按键序列， 字符 串获取装置 12根据所述输入按键序列来获取一个或多个候选字符串， 其具体过程与前述参照图 1所描述的实施例中输入获取装置 11和字符 串获取装置 12所执行的过程相同， 为简明起见， 以引用方式包含于此， 而不做赞述。
字符串提供装置将字符串获取装置 12 获取的一个或多个候选字符 串随机或按照一定的顺序， 如候选字符串的词频由高到低、 在候选词库 中的排列顺序等， 提供至该用户设备 1。 例如， 字符串获取装置 12获取 的候选字符串为" son"、 "sun", 才艮据这些候选字符串在候选词库中的词 频， 如" sun"的词频高于" son"的词频， 字符串提供装置将这些候选字符 串按词频由高到低的顺序提供给用户设备 1 , 即" sun"、 "son"。 优选地， 字符串提供装置将还可以将筛选装置与字符串获取装置 12 结合筛选获 取的一个或多个候选字符串按照上述方式提供至该用户设备 1。 本领域 技术人员应能理解上述提供候选字符串的方式仅为举例， 其他现有的或 今后可能出现的提供候选字符串的方式如可适用于本发明， 也应包含在 本发明保护范围以内， 并以引用方式包含于此。
在此， 字符串提供装置将候选字符串提供给用户设备 1 , 一方面， 用户可以通过该用户设备 1选择其所希望输入的候选字符串， 以用于获 取相应的候选词条； 另一方面， 该步骤还可以实现用户输入的输入按键 序列的扩展提醒， 进一步地， 在有些情况下， 候选字符串是对输入按键 序列的输入错误提醒， 可以使得用户及时了解其错误的输入习惯。
优选地， 用户设备 1还包括第一更新装置 （未示出） ， 该第一更新 装置根据该用户对所述一个或多个候选字符串的选择操作， 更新所述候 选词库。 具体地， 第一更新装置根据该用户对字符串提供装置提供的一 个或多个候选字符串的选择操作， 更新候选词库， 如更新用户选择的候 选字符串的词频、更新输入按键序列与候选字符串的映射关系等。例如， 假设字符串提供装置提供至用户设备 1 的候选字符串为" son"和" sun", 用户经由键盘或鼠标对这些候选字符串进行选择操作， 如该用户选择候 选字符串" son", 第一更新装置根据用户的该选择操作， 更新候选词库中 "son"的词频， 即将" son"的词频加 1。 又如， 输入按键序列为" sin", 字符 串提供装置提供至用户设备 1 的相应候选字符串为 "son"和" sun", 用户 选择候选字符串" son", 第一更新装置根据该选择操作， 更新候选词库中 输入按键序列与候选字符串的映射关系， 即" sin"和" son"。 本领域技术人 员应能理解上述更新候选词库的方式仅为举例， 其他现有的或今后可能 出现的更新候选词库的方式如可适用于本发明， 也应包含在本发明保护 范围以内， 并以引用方式包含于此。
在再一优选实施例中 （参照图 1 ) , 用户设备 1还包括词条获取装 置 （未示出）和词条提供装置 （未示出） ， 词条获取装置根据所述一个 或多个候选字符串， 在匹配词库中进行匹配查询， 以获得与所述一个或 多个候选字符串相对应的一个或多个候选词条； 词条提供装置将所述一 个或多个候选词条提供至用户设备 1。 以下参照图 1对该再一优选实施 例进行详细描述， 其中， 输入获取装置 11 获取的用户经由用户设备输 入的输入按键序列， 字符串获取装置 12根据所述输入按键序列来获取 一个或多个候选字符串， 其具体过程与前述参照图 1所描述的实施例中 输入获取装置 11和字符串获取装置 12所执行的过程相同，为简明起见， 以引用方式包含于此， 而不丈赞述。
词条获取装置根据字符串获取装置 12 获取的一个或多个候选字符 串， 在匹配词库中进行匹配查询， 以获得与这（些）候选字符串相对应 的一个或多个候选词条。 例如， 字符串获取装置 12 获取候选字符串 "sun", "son", 词条获取装置对这些候选字符串在匹配词库中进行匹配 查询， 获取相应的候选词条 "孙"、 "son"。 进一步地， 词条获取装置还可 以仅对与用户的选择操作相对应的候选字符串在匹配词库中进行匹配 查询， 以获取相应的候选词条； 假设用户选择候选字符串" sun", 则词条 获取装置仅对该候选字符串 "sun"在匹配词库中进行匹配查询，以获取相 应的候选词条 "孙"。 在此， 本领域技术人员应能理解， 匹配词库可以与 4矣选词库相互独立， 也可以合并为一个词库。
词条提供装置将词条获取装置获取的一个或多个候选词条提供至 该用户设备 1。 例如， 词条获取装置获取候选词条 "孙"和" son", 根据这 些候选词条在该候选词库中的用户使用频次， 如"孙"的用户使用频次高 于" son"的用户使用频次，词条提供装置将这些候选词条按用户使用频次 由高到低的顺序提供给用户设备 1 , 即"孙"、 "son"。 本领域技术人员应 能理解上述提供候选词条的方式仅为举例， 其他现有的或今后可能出现 的提供候选词条的方式如可适用于本发明， 也应包含在本发明保护范围 以内， 并以引用方式包含于此。
在此， 词条提供装置将词条获取装置获取的一个或多个候选词条提 供给用户设备 1 , 一方面， 供用户选择其希望输入的候选词条； 另一方 面， 该步骤还实现了对用户可选择的候选词条的扩展， 进一步地， 减少 了用户输入错误时的纠错过程， 同时也使得提高用户输入的准确性成为 可能。
优选地， 用户设备 1还包括匹配装置 （未示出） ， 该匹配装置才艮据 所述输入按键序列， 在所述匹配词库中进行匹配查询， 以获得与所述输 入按键序列相对应的一个或多个匹配词条； 词条提供装置将所述一个或 多个匹配词条与所述一个或多个候选词条提供至用户设备 1。 具体地， 匹配装置根据输入获取装置 11 获取的输入按键序列， 在该匹配词库中 进行匹配查询， 以获得与该输入按键序列相对应的一个或多个匹配词 条； 随后， 词条提供装置还将匹配装置获取的一个或多个匹配词条和词 条获取装置获取的候选词条提供至该用户设备 1 , 诸如将匹配词条和候 选词条合并， 进一步地， 还可以选择将词频最高的若干个匹配词条和词 频最高的若干个候选词条合并， 并按照合并后各词条的词频、 用户使用 频次等提供给用户设备 1。 例如， 假设输入按键序列为 "sin", 候选词库 和匹配词库为一个词库， 匹配装置对该输入按键序列在该词库中进行匹 配查询， 获得相应的匹配词条 "思念"、 "四年"、 "肆虐"、 "私念"， 字符 串获取装置 12获取相应的候选字符串， 如" son"和" sun", 词条获取装置 获得相应的候选词条 "son"、 "孙"， 词条提供装置根据这些匹配词条和候 选词条在该词库中的词频， 将这些匹配词条和候选词条按词频由高到低 的顺序提供给用户设备 1 , 如"思念"、 "孙"、 "四年"、 "肆虐"、 "son", "私 念"。本领域技术人员应能理解上述提供匹配词条和候选词条的方式仅为 举例， 其他现有的或今后可能出现的提供匹配词条和候选词条的方式如 可适用于本发明， 也应包含在本发明保护范围以内， 并以引用方式包含 于此。
在此， 词条提供装置将匹配词条和候选词条提供至该用户设备 1 , 一方面， 为用户提供更多的选择； 另一方面， 在匹配词条不是用户希望 输入的词条的情况下， 基于对输入按键序列的扩展提供候选词条， 减少 了用户需要重新输入的次数， 提高了用户的输入效率和输入准确性。
优选地， 用户设备 1还包括选择获取装置 （未示出） , 该选择获取 装置获取该用户对所述一个或多个候选字符串的选择操作； 词条获取装 置还根据所述选择操作所对应的候选字符串， 在所述匹配词库中进行匹 配查询， 以获得与所述选择操作所对应的候选字符串相对应的一个或多 个候选词条。 具体地， 选择获取装置获取该用户对一个或多个候选字符 串的选择操作； 词条获取装置还根据该选择操作所对应的候选字符串， 在该匹配词库中进行匹配查询， 以获得与该选择操作所对应的候选字符 串相对应的一个或多个候选词条。例如，假设候选字符串为" son"、 "sun", 选择获取装置获取用户对这些候选字符串的选择操作， 如该用户通过键 盘选择候选字符串 "sun"; 词条获取装置根据" sun"在匹配词库中进行匹 配查询， 获得候选词条 "孙"、 "损"、 "笋"等。 本领域技术人员应能理解 上述获取候选词条的方式仅为举例， 其他现有的或今后可能出现的获取 候选词条的方式如可适用于本发明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
更优选地， 用户设备 1还包括第二更新装置 （未示出） ， 该第二更 新装置根据该用户对所述一个或多个候选词条的选择操作， 更新所述匹 配词库。 具体地， 第二更新装置根据用户对候选词条的选择操作， 更新 匹配词库， 诸如更新用户选择的候选词条在该匹配词库中的词频、 更新 输入按键序列与用户选择的候选词条的映射关系等。 例如， 假设输入按 键序列 "sin", 字符串获取装置 12获取相应的候选字符串" son"和" sun", 词条获取装置获取相应的候选词条 "孙"、 "son", 词条提供装置将这两个 候选词条提供至用户设备 1 , 用户经由该用户设备 1选择候选词条 "孙"， 第二更新装置根据该选择操作，更新该匹配词库中"孙"的词频，即将"孙" 的词频加 1 , 同时更新输入按键序列" sin"与用户选择的候选词条"孙"的 映射关系。 本领域技术人员应能理解上述更新匹配词库的方式仅为举 例， 其他现有的或今后可能出现的更新匹配词库的方式如可适用于本发 明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
本领域技术人员应能理解， 在候选词库和匹配词库为一个词库的情 况下， 上述第一更新装置和第二更新装置可以为两个相互独立的模块， 也可以集成在一起。 图 4为根据本发明另一个方面的方法流程图， 示出获取与输入按 键序列相对应的候选字符串的过程。 具体地， 在步骤 S1中， 用户设备 1获取用户经由用户设备 1的键盘输入的输入按键序列； 随后， 在步骤 S2 中， 用户设备 1根据所述按键序列， 结合所述键盘的键盘布局， 在 候选词库中进行查询， 以获取与所述输入按键序列相对应的一个或多个 候选字符串。 在此， 用户设备 1可以是任何一种能与用户通过键盘进行 人机交互的电子产品， 例如计算机、 手机、 PDA、 平板电脑、 PSP、 掌 上电脑 PPC等； 其中， 键盘可以是任何一种物理键盘或虚拟键盘， 包括 但不限于 QWERTY全键盘、 普通手机键盘 ( keypad ) 、 按字母顺序排 列的全键盘， 以及任何自定义布局的虚拟键盘。 本领域技术人员应能理 解上述用户设备以及键盘仅为举例， 其他现有的或今后可能出现的用户 设备或者键盘如可适用于本发明， 也应包含在本发明保护范围以内， 并 以引用方式包含于此。
更具体地， 在步骤 S1 中， 用户设备 1获取用户经由用户设备 1的 键盘输入的输入按键序列。 具体地， 在步骤 S1 中， 用户设备 1通过检 测并解析用户按压用户设备 1的触摸屏上的虚拟键盘或者敲击物理键盘 等产生的电信号， 获取该用户经由该用户设备 1的键盘输入的输入按键 序列。 例如， 假设用户设备 1为带有触摸屏的电子产品， 相应的虚拟键 盘为 QWERTY全键盘， 用户通过按压该 QWERTY全键盘中的按键， 向该用户设备 1进行输入， 相应地， 用户设备 1通过检测并解析该用户 按压该触摸屏所产生的电信号， 获取该用户输入的输入按键序列。 本领 域技术人员应能理解上述获取输入按键序列的方式仅为举例， 其他现有 的或今后可能出现的获取输入按键序列的方式如可适用于本发明， 也应 包含在本发明保护范围以内， 并以引用方式包含于此。
随后， 在步骤 S2中， 用户设备 1根据所述输入按键序列， 结合所述 键盘的键盘布局， 在候选词库中进行查询， 以获取与所述输入按键序列 相对应的一个或多个候选字符串。 具体地， 在步骤 S2 中， 用户设备 1 根据其在步骤 S1中获取的输入按键序列， 结合用户设备 1的键盘布局， 在候选词库中进行查询， 以获取与该输入按键序列相对应的一个或多个 候选字符串。 在此， 用户设备 1获取候选字符串的方式包括但不限于： 1 )根据预设的基于键盘布局的字符映射表, 如" g→f、 h、 t、 v", "s→a、 d、 w、 x"等， 获取与输入按键序列相对应的一个或多个映射字符串， 如 根据" gongs"获得映射字符串" hongs"、 "fongs"、 "honha"、 "tonfd"等， 再 在候选词库中对这些映射字符进行匹配查询， 以获得候选字符串， 如 "hongs"; 2 )才艮据用户输入的输入按键序列， 在候选词库中进行查询， 以获取与输入按键序列相对应的一个或多个第一扩展字符串， 这些第一 扩展字符串所对应的按键与该输入按键序列的距离较近， 如根据输入按 键序列 "sin"获得第一扩展字符串" sun"、 "san"和" son", 再根据键盘布局， 从这些第一扩展字符串中选择候选字符串， 如 "u"、 "0"与" Γ按键位置相 邻， 则选择 "sun"和" son"作为候选字符串， 在此， "距离 "指第一扩展字符 串中字符在输入键盘中对应的按键与输入按键序列在该输入键盘中对 应的按键位置之间的距离； 3 )根据用户输入键盘的键盘布局， 获取与 输入按键序列相对应的一个或多个第二扩展字符串， 如 QWERTY全键 盘中" k"与" i"、 "ο"、 "1"、 "m"、 "j"按键位置相邻， "rkse"对应的第二扩 展字符串为" rise"、 "rose", "rise", "rmse"、 "rjse", 再在候选词库中对这 些第二扩展字符串进行匹配查询， 获得候选字符串， 如" rise"和" rose"。
需要说明的是， 本领域技术人员应能理解， 上述 3种获取候选字符 串的方式之间不是互相排斥的， 而是可以共同在一次获取候选字符串的 过程中实现的。 例如， 用户设备 1根据用户的输入按键序列， 分别按照 上述 3种获取候选字符串的方式获取相应的候选字符串， 并将 3种方式 所获得的候选字符串合并， 以获得最终与用户的输入按键序列相对应的 候选字符串。
本领域技术人员还应能理解上述获取候选字符串的方式仅为举例， 其他现有的或今后可能出现的获取候选字符串的方式如可适用于本发 明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。优选地， 候选词库包括本地词库和网络词库； 并且， 候选词库还包括中文词库、 英文词库、 其他语言词库， 以及多种语言混合的词库。
优选地， 用户设备 1在步骤 S1和步骤 S2中是持续不断地工作。 具 体地， 在步骤 S1 中， 用户设备 1用于获取用户经由用户设备 1的键盘 输入的输入按键序列； 在步骤 S2中， 用户设备 1用于根据所述输入按 键序列， 结合所述键盘的键盘布局， 在候选词库中进行查询， 以获取与 所述输入按键序列相对应的一个或多个候选字符串。 在此， 本领域技术 人员应理解"持续"是指用户设备 1在步骤 S1和步骤 S2中分别按照设定 的或实时调整的工作模式要求进行输入按键序列的获取、 以及候选字符 串的获取， 直至用户在较长时间内停止经由用户设备 1的键盘进行输入 按键序列的输入。
图 5为根据本发明一个优选实施例的方法流程图， 示出获取与输 入按键序列相对应的候选字符串的过程。 具体地， 在步骤 S21，中， 用 户设备 1根据所述输入按键序列， 在所述候选词库中进行查询， 以获取 与所述输入按键序列相对应的一个或多个第一扩展字符串； 在步骤 S22' 中， 用户设备 1根据所述键盘布局， 从所述一个或多个第一扩展字符串 中选择所述一个或多个候选字符串。
更具体地， 在步骤 S21，中， 用户设备 1根据所述输入按键序列， 在 所述候选词库中进行查询， 以获取与所述输入按键序列相对应的一个或 多个第一扩展字符串。 具体地， 在步骤 S21，中， 用户设备 1才艮据用户经 由用户设备 1的键盘输入的输入按键序列， 在候选词库中进行查询， 以 获取与该输入按键序列相对应的一个或多个第一扩展字符串。 例如， 假 设输入按键序列" sin"， 候选词库为用户设备 1的本地词库， 用户设备 1 在该候选词库中对" sm"进行扩展查询， 得到相应的第一扩展字符串 "sun", "san"和" son"。 本领域技术人员应能理解上述获取第一扩展字符 串的方式仅为举例， 其他现有的或今后可能出现的获取第一扩展字符串 的方式如可适用于本发明， 也应包含在本发明保护范围以内， 并以引用 方式包含于此。
随后， 在步骤 S22，中， 用户设备 1根据所述键盘布局， 从所述一个 或多个第一扩展字符串中选择所述一个或多个候选字符串。 具体地， 在 步骤 S22，中，用户设备 1根据用户输入键盘的键盘布局，从其在步骤 S21 ' 中获取的一个或多个第一扩展字符串中选择与该键盘布局相对应的一 个或多个候选字符串。 例如， 接上例， 用户设备 1在步骤 S21，中获得第 一扩展字符串 "sun"、 "san" 和" son", 用户的输入键盘为 QWERTY全键 盘， "u"、 "0"与" i"按键位置相邻， 在步骤 S22，中， 用户设备 1从这些第 一扩展字符串中选择 "sun"和" son"作为候选字符串。 本领域技术人员应 能理解上述获取候选字符串的方式仅为举例， 其他现有的或今后可能出 现的获取候选字符串的方式如可适用于本发明， 也应包含在本发明保护 范围以内， 并以引用方式包含于此。
图 6为根据本发明另一优选实施例的方法流程图， 示出获取与输 入按键序列相对应的候选字符串的过程。 具体地， 在步骤 S23"中， 用 户设备 1 根据所述输入按键序列， 结合所述键盘布局， 获取与所述输入 按键序列相对应的一个或多个第二扩展字符串； 在步骤 S24"中， 用户设 备 1 在所述候选词库中对所述一个或多个第二扩展字符串进行匹配查 询， 以获取所述一个或多个候选字符串。
更具体地， 在步骤 S23"中， 用户设备 1根据所述输入按键序列， 结合所述键盘布局， 获取与该输入按键序列相对应的一个或多个第二扩 展字符串。 具体地， 在步骤 S23"中， 用户设备 1根据用户经由用户设 备 1输入的输入按键序列， 结合该用户设备 1的键盘布局， 获取与该输 入按键序列相对应的一个或多个第二扩展字符串。 例如， 假设输入按键 序列" uing", 用户设备 1的键盘为 QWERTY全键盘， 根据该键盘布局， "u"与" y"、 "Γ、 "J"按键位置相邻， 据此， 用户设备 1获得第二扩展字符 串" ying"、 "iing"、 "jing"。 又如， 假设输入按键序列" jent", 用户设备 1 为手机， 其键盘为普通手机键盘 （keypad ) , 即一个按键对应多个字母 的键盘， 根据该键盘布局， 按键" ghi"、 按键" mno" 与按键 "jkl"按键位 置相邻， 据此， 用户设备 1获得第二扩展字符串 "gent"、 "hent"、 "ient"、 "ment"、 "nent"、 "oent"、 "kent"、 "lent'O 本领域技术人员应能理解上述 获取第二扩展字符串的方式仅为举例， 其他现有的或今后可能出现的获 取第二扩展字符串的方式如可适用于本发明， 也应包含在本发明保护范 围以内， 并以引用方式包含于此。
随后， 在步骤 S24"中， 用户设备 1在所述候选词库中对所述一个 或多个第二扩展字符串进行匹配查询， 以获取所述一个或多个候选字符 串。 具体地， 在步骤 S24"中， 用户设备 1 在该候选词库中对其在步骤 S23"中获得的一个或多个第二扩展字符串进行匹配查询， 以获取一个或 多个候选字符串。 例如，接上例， 用户设备 1在步骤 S23"中获得第二扩 展字符串" ying"、 "iing"、 "jing", 在步骤 S24"中， 用户设备 1在候选词 库中对这些第二扩展字符串进行匹配查询， 获得候选字符串 "ymg"和 "jing"。 又如， 用户设备 1 在步骤 S23"中获得第二扩展字符串" gent"、 "hent"、 "ient"、 "ment"、 "nent"、 "oent"、 "kent"、 "lent", 在步骤 S24" 中， 用户设备 1在候选词库中对这些第二扩展字符串进行匹配查询， 得 到候选字符串 "gent"、 "hent"、 "kent"、 "lent"。 本领域技术人员应能理解 上述获取候选字符串的方式仅为举例， 其他现有的或今后可能出现的获 取候选字符串的方式如可适用于本发明， 也应包含在本发明保护范围以 内， 并以引用方式包含于此。
在一个优选实施例中（参照图 4 ) ,该过程还包括步骤 S3 (未示出）， 在步骤 S3中， 用户设备 1根据查询辅助信息， 对所述一个或多个候选 字符串进行筛选， 以获取筛选后的一个或多个候选字符串。 以下参照图 4对该优选实施例进行详细描述， 其中， 在步骤 S1中， 用户设备 1用于 获取用户经由用户设备 1 的键盘输入的输入按键序列； 在步骤 S2中， 用户设备 1用于根据所述输入按键序列， 结合所述键盘的键盘布局， 在 候选词库中进行查询， 以获取与所述输入按键序列相对应的一个或多个 候选字符串； 其具体过程与前述参照图 4所描述的实施例中用户设备 1 在步骤 S1中和在步骤 S2中所执行的过程相同， 为简明起见， 以引用方 式包含于此， 而不做赞述。
具体地， 在步骤 S3中， 用户设备 1才艮据查询辅助信息， 诸如该用户 设备 1 的当前应用、 该用户的用户特征信息等， 对其在步骤 S2中获取 的一个或多个候选字符串进行筛选， 或者与步骤 S2相结合， 同时获取 及筛选候选字符串， 以获得筛选后的一个或多个候选字符串。 例如， 在 步骤 S2中，用户设备 1根据输入按键序列 "sm"获得的候选字符串为 "son" 和" sun", 假设该用户设备 1的当前应用为手机电话薄的联系人输入， 且 为中文输入， 在步骤 S3中， 用户设备 1根据该当前应用对候选字符串 "son"和" sun"进行筛选， 得到筛选后的候选字符串" sun"。 又如， 输入按 键序列为" sin" , 用户的输入键盘为 QWERTY全键盘， 用户设备 1根据 该用户之前的输入历史记录 "son",以及该输入按键序列 "sin"和该输入键 盘的键盘布局， 如" i"的相邻按键为 "u"和" 0" , 在候选词库中进行查询， 得到候选字符串" son"。本领域技术人员应能理解上述根据查询辅助信息 筛选候选字符串的方式仅为举例， 其他现有的或今后可能出现的根据查 询辅助信息筛选候选字符串的方式如可适用于本发明， 也应包含在本发 明保护范围以内， 并以引用方式包含于此。
优选地， 查询辅助信息包括但不限于， 以下至少任一项：
1) 用户设备 1的当前应用；
2) 该用户的用户特征信息；
3) 所述输入按键序列的在先输入信息；
4) 所述候选字符串所对应的语义分析结果；
5) 所述候选字符串与所述输入按键序列的距离阈值。
具体地， 1 )用户设备 1的当前应用指用户设备 1当前运行的应用程 序。 在步骤 S3中， 用户设备 1根据其当前运行的应用程序， 对其在步 骤 S2 中获取的一个或多个候选字符串进行筛选， 或者根据其当前运行 的应用程序， 以及用户输入的输入按键序列和输入键盘的键盘布局， 在 候选词库中进行匹配查询， 筛选获得候选字符串， 以使这（些）候选字 符串不仅满足基于键盘布局的对输入按键序列的扩展， 而且符合该用户 设备 1的当前应用。 例如， 用户输入的输入按键序列为 "sin" , 用户设备 1的当前应用为电话簿的联系人输入，用户设备 1在步骤 S2中获取相应 的候选字符串为" son"、 "sun", 在步骤 S3中， 该用户设备 1才艮据该当前 应用对这些候选字符串进行筛选， 获得候选字符串" sun"。
2 )用户的用户特征信息包括用户的输入历史记录、用户的输入偏好 设置等， 在步骤 S3中， 用户设备 1根据用户的用户特征信息， 诸如用 户的输入历史记录、 用户的输入偏好设置等， 对其在步骤 S2 中获取的 一个或多个候选字符串进行筛选， 或者与步骤 S2相结合， 根据用户的 用户特征信息， 以及用户输入的输入按键序列和输入键盘的键盘布局， 在候选词库中进行匹配查询， 筛选获得候选字符串， 以使这（些）候选 字符串不仅满足基于键盘布局的对输入按键序列的扩展， 而且符合该用 户的用户特征信息。 例如， 用户的输入偏好设置为" m→n", 即该用户习 惯性的在敲击 "n"时误敲击" m", 假设输入按键序列为 "ming", 输入键盘 为按字母顺序排列的全键盘， 用户设备 1根据该输入偏好设置， 以及该 输入按键序列和该键盘布局， 如" m"的相邻按键为 "n"和" Γ , 确定对 "ming"进行扩展的字符串为 "ning", 并对该 "ning"在候选词库中进行查 询， 获得候选字符串" nmg"。
3 )输入按键序列的在先输入信息包括用户设备 1的緩存中记录的在 该输入按键序列之前输入的字符串和词条， 在步骤 S3 中， 用户设备 1 根据这（些）在先输入信息与其在步骤 S2 中获取的候选字符串之间的 上下文语义分析， 对这（些）候选字符串进行筛选， 或者与步骤 S2相 结合， 根据这（些）在先输入信息， 以及用户输入的输入按键序列和输 入键盘的键盘布局， 在候选词库中进行匹配查询和上下文语义分析， 筛 选获得候选字符串， 以使这（些）候选字符串不仅满足基于键盘布局的 对输入按键序列的扩展， 而且符合在先输入信息的语义。 例如， 用户输 入的输入按键序列为 "ahiwuduo", 用户设备 1在步骤 S2中获取相应的 候选字符串， 如" shiwusuo"、 "zhisuduo"、 "shiwuduo"、 "zhiwusuo", 用 户设备 1的緩存中记录该用户在输入该输入按键序列之前输入的词条为 "律师"，在步骤 S3中， 用户设备 1对该在先输入信息分别与上述候选字 符串进行上下文语义分析， 确定候选字符串为 "shiwusuo"。 4 )候选字符串所对应的语义分析结果包括通过对候选字符串本身的 语义分析所获得的结果， 如候选字符串是否有意义， 在步骤 S3 中， 用 户设备 1对其在步骤 S2中获取的候选字符串做语义分析， 根据这（些） 语义分析结果对候选字符串进行筛选， 或者与步骤 S2相结合， 根据用 户输入的输入按键序列和输入键盘的键盘布局， 在候选词库中进行匹配 查询， 同时对候选字符串做语义分析， 根据该语义分析结果， 筛选获得 候选字符串， 以使这（些）候选字符串不仅满足基于键盘布局的对输入 按键序列的扩展， 而且候选字符串本身有意义。 例如， 假设用户输入的 输入按键序列为" sime people", 该用户的输入键盘为 QWERTY全键盘， 用户设备 1在步骤 S2中基于该输入按键序列和键盘布局， 如" i"的相邻 按键为 "o"、 "u"等," s"的相邻按键为 "d"、 "w"、 "x"等， 在候选词库中进 行查询 , 获得候选字符串" some people"、 "dime people"和" dome people", 在步骤 S3中， 用户设备 1分别对这些候选字符串做语义分析， 并根据 相应的语义分析结果确定候选字符串， 如" some people"
5 )候选字符串与该输入按键序列的距离阈值包括候选字符串与输入 按键序列之间不相同字符数的预设最大值， 以及候选字符串中字符在输 入键盘中对应的按键位置与输入按键序列在该输入键盘中对应的按键 位置之间的预设最大距离。 在步骤 S3中， 用户设备 1按照该距离阈值， 对其在步骤 S2中获取的候选字符串进行筛选， 或者与步骤 S2相结合， 根据该距离阈值， 以及用户输入的输入按键序列和输入键盘的键盘布 局， 在候选词库中进行匹配查询， 筛选获得候选字符串， 以使这（些） 候选字符串不仅满足基于键盘布局的对输入按键序列的扩展， 而且不超 过候选字符串与该输入按键序列的距离阈值。 例如， 用户输入按键序列 为" ahurufa", 假设预设距离阈值为 1 , 在此， 距离阈值指候选字符串与 输入按键序列之间不相同字符数的预设最大值， 即候选字符串最多只能 与该输入按键序列有 1个字符不同， 用户的输入键盘为 QWERTY全键 盘， 用户设备 1按照该距离阈值， 以及用户输入的输入按键序列和输入 键盘的键盘布局，如" a"的相邻按键为 "z"、 "s"等," u"的相邻按键为 "i"、 "y" 等，在候选词库中进行查询，筛选得到候选字符串为 "shurufs"、 "zhurufa"。 又如， 用户输入按键序列为 "jind", 假设距离阈值为 1 , 在此， 距离阈值 指候选字符串中字符在输入键盘中对应的按键位置与输入按键序列在 该输入键盘中对应的按键位置之间的预设最大距离， 输入键盘为 QWERTY全键盘， 用户设备 1按照该距离阈值， 以及用户输入的输入 按键序列和输入键盘的键盘布局， 确定" j"的相邻按键为 "k"、 "m"、 "h" 等," Γ的相邻按键为 "u"等， 并在候选词库中进行查询， 得到候选字符串 "kind", "mind", "hind", "jund"。
本领域技术人员应能理解， 上述各项查询辅助信息不仅可以单独用 于筛选候选字符串， 还可以两两结合、 甚至相互结合用于筛选候选字符 串。 本领域技术人员还应能理解上述各项查询辅助信息仅为举例， 其他 现有的或今后可能出现的查询辅助信息如可适用于本发明， 也应包含在 本发明保护范围以内， 并以引用方式包含于此。
更优选地， 用户特征信息包括但不限于， 以下至少任一项：
1 )该用户的输入历史记录；
2 )该用户的输入偏好设置。
具体地， 1 )用户的输入历史记录， 包括但不限于用户历史输入的输 入按键序列， 用户对候选字符串的选择操作， 用户历史输入的输入按键 序列与用户选择的候选字符串的映射关系等， 如输入按键序列为 "sm" , 相应的候选字符串为 "sun"和" son", 用户选择候选字符串 "sun", 据此用 户设备 1建立" sin"与 "sun"的映射关系。 2 )用户的输入偏好设置， 包括 但不限于用户根据其输入习惯而特别设置的字符之间的映射关系， 如 "m→n", 以及用户对输入法的偏好， 如英文输入的优先级高于中文输入 的优先级。 本领域技术人员应能理解， 上述各项用户特征信息不仅可以 单独用于筛选候选字符串， 还可以两两结合、 甚至相互结合用于筛选候 选字符串。 本领域技术人员应能理解上述各项用户特征信息仅为举例， 其他现有的或今后可能出现的用户特征信息如可适用于本发明 , 也应包 含在本发明保护范围以内， 并以引用方式包含于此。 在另一优选实施例中（参照图 4 ) ,该过程还包括步骤 S4 (未示出）， 在步骤 S4中， 用户设备 1将所述一个或多个候选字符串提供至用户设 备 1。 以下参照图 4对该另一优选实施例进行详细描述， 其中， 在步骤 S1中，用户设备 1用于获取用户经由用户设备 1的键盘输入的输入按键 序列； 在步骤 S2中， 用户设备 1用于根据所述输入按键序列， 结合所 述键盘的键盘布局， 在候选词库中进行查询， 以获取与所述输入按键序 列相对应的一个或多个候选字符串； 其具体过程与前述参照图 4所描述 的实施例中用户设备 1在步骤 S1中和在步驟 S2中所执行的过程相同， 为简明起见， 以引用方式包含于此， 而不做赞述。
在步骤 S4中， 用户设备 1将其在步骤 S2中获取的一个或多个候选 字符串随机或按照一定的顺序， 如候选字符串的词频由高到低、 在候选 词库中的排列顺序等， 提供至该用户设备 1。 例如， 在步骤 S2中， 用户 设备 1 获取的候选字符串为 "son"、 "sun", 根据这些候选字符串在候选 词库中的词频， 如" sun"的词频高于" son"的词频， 在步骤 S4中， 用户设 备 1将这些候选字符串按词频由高到低的顺序提供给该用户设备 1 , 即 "sun", "son"。 优选地， 在步骤 S4中， 用户设备 1将还可以将其在步骤 S3 中筛选获取的一个或多个候选字符串按照上述方式提供至该用户设 备 1。 本领域技术人员应能理解上述提供候选字符串的方式仅为举例， 其他现有的或今后可能出现的提供候选字符串的方式如可适用于本发 明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
在此， 在步骤 S4中， 用户设备 1提供候选字符串， 一方面， 用户 可以通过该用户设备 1选择其所希望输入的候选字符串， 以用于获取相 应的候选词条； 另一方面， 该步骤还可以实现用户输入的输入按键序列 的扩展提醒， 进一步地， 在有些情况下， 候选字符串是对输入按键序列 的输入错误提醒， 可以使得用户及时了解其错误的输入习惯。
优选地， 该过程还包括步骤 S5 (未示出） ， 在步骤 S5中， 用户设 备 1根据该用户对所述一个或多个候选字符串的选择操作， 更新所述候 选词库。 具体地， 在步骤 S5中， 用户设备 1根据该用户对其在步骤 S4 中提供的一个或多个候选字符串的选择操作， 更新候选词库， 如更新用 户选择的候选字符串的词频、 更新输入按键序列与候选字符串的映射关 系等。 例如， 在步骤 S4中， 用户设备 1提供候选字符串" son"和" sun", 用户经由键盘或鼠标对这些候选字符串进行选择操作， 如该用户选择候 选字符串" son", 在步骤 S5中， 用户设备 1根据用户的该选择操作， 更 新候选词库中 "son"的词频， 即将" son"的词频加 1。 又如， 输入按键序列 为" sin", 在步骤 S4 中， 用户设备 1 提供的相应候选字符串为" son"和 "sun", 用户选择候选字符串 "son", 在步骤 S5中， 用户设备 1根据该 选择操作， 更新候选词库中输入按键序列与候选字符串的映射关系， 即 "sin"和" son"。本领域技术人员应能理解上述更新候选词库的方式仅为举 例， 其他现有的或今后可能出现的更新候选词库的方式如可适用于本发 明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
在再一优选实施例中（参照图 4 ) , 该过程还包括步骤 S6 (未示出） 和 S7 (未示出） ， 在步骤 S6中， 用户设备 1根据所述一个或多个候选 字符串， 在匹配词库中进行匹配查询， 以获得与所述一个或多个候选字 符串相对应的一个或多个候选词条； 在步骤 S7中， 用户设备 1将所述 一个或多个候选词条提供至用户设备 1。 以下参照图 4对该再一优选实 施例进行详细描述， 其中， 在步骤 S1 中， 用户设备 1用于获取用户经 由用户设备 1 的键盘输入的输入按键序列； 在步骤 S2中， 用户设备 1 用于根据所述输入按键序列， 结合所述键盘的键盘布局， 在候选词库中 进行查询， 以获取与所述输入按键序列相对应的一个或多个候选字符 串；其具体过程与前述参照图 4所描述的实施例中用户设备 1在步骤 S1 中和在步骤 S2 中所执行的过程相同， 为简明起见， 以引用方式包含于 此， 而不^ ί故赞述。
在步骤 S6中， 用户设备 1根据其在步骤 S2中获取的一个或多个候 选字符串， 在匹配词库中进行匹配查询， 以获得与这（些）候选字符串 相对应的一个或多个候选词条。 例如， 用户设备 1在步骤 S2中获取候 选字符串 "sun"、 "son", 在步骤 S6中， 用户设备 1对这些候选字符串 在匹配词库中进行匹配查询， 获取相应的候选词条"孙"、 "son"。 进一步 地， 在步骤 S6中， 用户设备 1还可以仅对与用户的选择操作相对应的 候选字符串在匹配词库中进行匹配查询， 以获取相应的候选词条； 假设 用户选择候选字符串" sun" , 则词条获取装置仅对该候选字符串 "sun"在 匹配词库中进行匹配查询， 以获取相应的候选词条"孙"。 在此， 本领域 技术人员应能理解， 匹配词库可以与候选词库相互独立， 也可以合并为 一个词库。
在步骤 S7中， 用户设备 1将其在步骤 S6中获取的一个或多个候选 词条提供至该用户设备 1。 例如， 用户设备 1在步骤 S6中获取候选词条 "孙"和" son" , 根据这些候选词条在该候选词库中的用户使用频次， 如 "孙"的用户使用频次高于 "son"的用户使用频次， 在步骤 S7 中， 用户设 备 1将这些候选词条按用户使用频次由高到低的顺序提供给用户设备 1 , 即"孙"、 "son"。 本领域技术人员应能理解上述提供候选词条的方式仅为 举例， 其他现有的或今后可能出现的提供候选词条的方式如可适用于本 发明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
在此， 用户设备 1提供其获取的一个或多个候选词条， 一方面， 供 用户选择其希望输入的候选词条； 另一方面， 还实现了对用户可选择的 候选词条的扩展， 进一步地， 减少了用户输入错误时的纠错过程， 同时 也使得提高用户输入的准确性成为可能。
优选地， 该过程还包括步骤 S8 (未示出） ， 在步骤 S8中， 用户设 备 1根据所述输入按键序列， 在所述匹配词库中进行匹配查询， 以获得 与所述输入按键序列相对应的一个或多个匹配词条； 在步骤 S7 中， 用 户设备 1将所述一个或多个匹配词条与所述一个或多个候选词条提供至 用户设备 1。 具体地， 在步骤 S8中， 用户设备 1根据其在步骤 S1中获 取的输入按键序列， 在该匹配词库中进行匹配查询， 以获得与该输入按 键序列相对应的一个或多个匹配词条； 随后， 在步骤 S7中， 用户设备 1 还将其在步骤 S8中获取的一个或多个匹配词条和其在步骤 S6中获取的 候选词条提供至该用户设备 1 , 诸如将匹配词条和候选词条合并， 进一 步地， 还可以选择将词频最高的若干个匹配词条和词频最高的若干个候 选词条合并， 并按照合并后各词条的词频、 用户使用频次等提供给用户 设备 1。 例如， 假设输入按键序列为 "sin", 候选词库和匹配词库为一个 词库， 用户设备 1对该输入按键序列在该词库中进行匹配查询， 获得相 应的匹配词条 "思念"、 "四年"、 "肆虐"、 "私念"， 并获取相应的候选字 符串， 如" son"和" sun", 从而获得相应的候选词条" son"、 "孙 "， 用户设 备 1根据这些匹配词条和候选词条在该词库中的词频， 将这些匹配词条 和候选词条按词频由高到低的顺序提供给用户设备 1 , 如"思念"、 "孙"、 "四年"、 "肆虐"、 "son", "私念"。 本领域技术人员应能理解上述提供匹 配词条和候选词条的方式仅为举例， 其他现有的或今后可能出现的提供 匹配词条和候选词条的方式如可适用于本发明， 也应包含在本发明保护 范围以内， 并以引用方式包含于此。
在此， 用户设备 1将匹配词条和候选词条提供， 一方面， 为用户提 供更多的选择； 另一方面， 在匹配词条不是用户希望输入的词条的情况 下， 基于对输入按键序列的扩展提供候选词条， 减少了用户需要重新输 入的次数， 提高了用户的输入效率和输入准确性。
优选地， 该过程还包括步骤 S9 (未示出） ， 在步骤 S9中， 用户设 备 1 获取该用户对所述一个或多个候选字符串的选择操作； 在步骤 S6 中， 用户设备 1还根据所述选择操作所对应的候选字符串， 在所述匹配 词库中进行匹配查询， 以获得与所述选择操作所对应的候选字符串相对 应的一个或多个候选词条。 具体地， 在步骤 S9中， 用户设备 1获取该 用户对一个或多个候选字符串的选择操作； 在步骤 S6 中， 用户设备 1 还根据该选择操作所对应的候选字符串， 在该匹配词库中进行匹配查 询， 以获得与该选择操作所对应的候选字符串相对应的一个或多个候选 词条。 例如， 假设候选字符串为" son"、 "sun", 用户设备 1 获取用户对 这些候选字符串的选择操作， 如该用户通过键盘选择候选字符串 "sun"; 用户设备 1根据" sun"在匹配词库中进行匹配查询， 获得候选词条"孙"、 "损"、 "笋"等。 本领域技术人员应能理解上述获取候选词条的方式仅为 举例， 其他现有的或今后可能出现的获取候选词条的方式如可适用于本 发明， 也应包含在本发明保护范围以内， 并以引用方式包含于此。
更优选地， 该过程还包括步骤 S10 (未示出） ， 在步骤 S10中， 用 户设备 1根据该用户对所述一个或多个候选词条的选择操作， 更新所述 匹配词库。 具体地， 在步骤 S10中， 用户设备 1才艮据用户对候选词条的 选择操作， 更新匹配词库， 诸如更新用户选择的候选词条在该匹配词库 中的词频、 更新输入按键序列与用户选择的候选词条的映射关系等。 例 如， 假设输入按键序列" sin", 用户设备 1 获取相应的候选字符串" son" 和" sun", 以及相应的候选词条 "孙"、 "son", 将这两个候选词条提供至用 户设备 1 , 用户经由该用户设备 1选择候选词条"孙"， 用户设备 1根据 该选择操作， 更新该匹配词库中"孙"的词频， 即将"孙"的词频加 1 , 同 时更新输入按键序列 "sin"与用户选择的候选词条"孙"的映射关系。 本领 域技术人员应能理解上述更新匹配词库的方式仅为举例， 其他现有的或 今后可能出现的更新匹配词库的方式如可适用于本发明， 也应包含在本 发明保护范围以内， 并以引用方式包含于此。
对于本领域技术人员而言， 显然本发明不限于上述示范性实施例的 细节， 而且在不背离本发明的精神或基本特征的情况下， 能够以其他的 具体形式实现本发明。 因此， 无论从哪一点来看， 均应将实施例看作是 示范性的， 而且是非限制性的， 本发明的范围由所附权利要求而不是上 述说明限定， 因此旨在将落在权利要求的等同要件的含义和范围内的所 有变化涵括在本发明内。 不应将权利要求中的任何附图标记视为限制所 涉及的权利要求。 此外， 显然"包括"一词不排除其他单元或步骤， 单数 不排除复数。 装置权利要求中陈述的多个单元或装置也可以由一个单元 或装置通过软件或者硬件来实现。 第一， 第二等词语用来表示名称， 而 并不表示任何特定的顺序。

Claims

权 利 要 求 书
1. 一种获取与输入按键序列相对应的候选字符串的方法， 其中， 该 方法包括以下步骤：
a获取用户经由用户设备的键盘输入的输入按键序列；
b根据所述输入按键序列， 结合所述键盘的键盘布局， 在候选词库 中进行查询， 以获取与所述输入按键序列相对应的一个或多个候选字符 串。
2. 根据权利要求 1所述的方法， 其中， 所述步骤 b包括：
-根据所述输入按键序列， 在所述候选词库中进行查询， 以获取与 所述输入按键序列相对应的一个或多个第一扩展字符串；
-根据所述键盘布局， 从所述一个或多个第一扩展字符串中选择所 述一个或多个候选字符串。
3. 根据权利要求 1或 2所述的方法， 其中， 所述步骤 b包括： -根据所述输入按键序列， 结合所述键盘布局， 获取与所述输入按 键序列相对应的一个或多个第二扩展字符串；
- 在所述候选词库中对所述一个或多个第二扩展字符串进行匹配查 询， 以获取所述一个或多个候选字符串。
4. 根据权利要求 1至 3中任一项所述的方法，其中，该方法还包括： -根据查询辅助信息， 对所述一个或多个候选字符串进行筛选， 以 获取筛选后的一个或多个候选字符串。
5. 根据权利要求 4所述的方法 ， 其中， 所述查询辅助信息包括以 下至少任一项：
- 所述用户设备的当前应用；
- 该用户的用户特征信息；
所述输入按键序列的在先输入信息；
- 所述候选字符串所对应的语义分析结果； - 所述候选字符串与所述输入按键序列的距离阈值。
6. 根据权利要求 5所述的方法， 其中， 所述用户特征信息包括以下 至少任一项：
- 该用户的输入历史记录；
- 该用户的输入偏好设置。
7. 根据权利要求 1至 6中任一项所述的方法，其中，该方法还包括： - 将所述一个或多个候选字符串提供至所述用户设备。
8. 根据权利要求 7所述的方法， 其中， 该方法还包括：
-根据该用户对所述一个或多个候选字符串的选择操作， 更新所述 候选词库。
9. 根据权利要求 1至 8 中任一项所述的方法， 其中，该方法还包括： X才艮据所述一个或多个候选字符串， 在匹配词库中进行匹配查询， 以获得与所述一个或多个候选字符串相对应的一个或多个候选词条； y将所述一个或多个候选词条提供至所述用户设备。
10. 根据权利要求 9所述的方法， 其中， 该方法还包括：
-才艮据所述输入按键序列， 在所述匹配词库中进行匹配查询， 以获 得与所述输入按键序列相对应的一个或多个匹配词条；
其中， 所述步骤 y包括：
- 将所述一个或多个匹配词条与所述一个或多个候选词条提供至所 述用户设备。
11. 根据权利要求 9或 10所述的方法， 其中， 该方法还包括： - 获取该用户对所述一个或多个候选字符串的选择操作；
其中， 所述步骤 X包括：
-根据所述选择操作所对应的候选字符串， 在所述匹配词库中进行 匹配查询， 以获得与所述选择操作所对应的候选字符串相对应的一个或 多个候选词条。
12. 根据权利要求 9至 11中任一项所述的方法， 其中， 该方法还包 括： -根据该用户对所述一个或多个候选词条的选择操作， 更新所述匹 配词库。
13. 一种获取与输入按键序列相对应的候选字符串的设备， 其中， 该设备包括：
输入获取装置， 用于获取用户经由用户设备的键盘输入的输入按键 序列；
字符串获取装置， 用于根据所述输入按键序列， 结合所述键盘的键 盘布局， 在候选词库中进行查询， 以获取与所述输入按键序列相对应的 一个或多个候选字符串。
14. 根据权利要求 13所述的设备，其中，所述字符串获取装置包括： 第一扩展单元， 用于根据所述输入按键序列， 在所述候选词库中进 行查询， 以获取与所述输入按键序列相对应的一个或多个第一扩展字符 串；
选择单元， 用于根据所述键盘布局， 从所述一个或多个第一扩展字 符串中选择所述一个或多个候选字符串。
15. 根据权利要求 13或 14所述的设备， 其中， 所述字符串获取装 置包括：
第二扩展单元， 用于根据所述输入按键序列， 结合所述键盘布局， 获取与所述输入按键序列相对应的一个或多个第二扩展字符串；
查询单元， 用于在所述候选词库中对所述一个或多个第二扩展字符 串进行匹配查询， 以获取所述一个或多个候选字符串。
16. 根据权利要求 13至 15中任一项所述的设备， 其中， 该设备还 包括：
筛选装置， 用于才艮据查询辅助信息， 对所述一个或多个候选字符串 进行筛选， 以获取筛选后的一个或多个候选字符串。
17. 根据权利要求 16所述的设备， 其中， 所述查询辅助信息包括以 下至少任一项：
- 所述用户设备的当前应用； - 该用户的用户特征信息；
所述输入按键序列的在先输入信息；
- 所述候选字符串所对应的语义分析结果；
- 所述候选字符串与所述输入按键序列的距离阈值。
18. 根据权利要求 17所述的设备， 其中， 所述用户特征信息包括以 下至少任一项：
- 该用户的输入历史记录；
- 该用户的输入偏好设置。
19. 根据权利要求 13至 18中任一项所述的设备， 其中， 该设备还 包括：
字符串提供装置， 用于将所述一个或多个候选字符串提供至所述用 户设备。
20. 根据权利要求 19所述的设备， 其中， 该设备还包括：
第一更新装置， 用于才艮据该用户对所述一个或多个候选字符串的选 择操作， 更新所述候选词库。
21. 根据权利要求 13至 20 中任一项所述的设备， 其中， 该设备还 包括：
词条获取装置， 用于才艮据所述一个或多个候选字符串， 在匹配词库 中进行匹配查询， 以获得与所述一个或多个候选字符串相对应的一个或 多个候选词条；
词条提供装置， 用于将所述一个或多个候选词条提供至所述用户设 备。
22. 根据权利要求 21所述的设备， 其中， 该设备还包括：
匹配装置， 用于根据所述输入按键序列， 在所述匹配词库中进行匹 配查询， 以获得与所述输入按键序列相对应的一个或多个匹配词条； 其中， 所述词条提供装置用于：
- 将所述一个或多个匹配词条与所述一个或多个候选词条提供至所 述用户设备。
23. 根据权利要求 21或 22所述的设备， 其中， 该设备还包括： 选择获取装置， 用于获取该用户对所述一个或多个候选字符串的选 择操作；
其中， 所述词条获取装置用于：
-根据所述选择操作所对应的候选字符串， 在所述匹配词库中进行 匹配查询， 以获得与所述选择操作所对应的候选字符串相对应的一个或 多个候选词条。
24. 根据权利要求 21至 23中任一项所述的设备， 其中， 该设备还 包括：
第二更新装置， 用于才艮据该用户对所述一个或多个候选词条的选择 操作， 更新所述匹配词库。
PCT/CN2012/073589 2011-04-11 2012-04-06 获取与输入按键序列相对应的候选字符串的方法与设备 WO2012139475A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201110089966.3 2011-04-11
CN2011100899663A CN102184028A (zh) 2011-04-11 2011-04-11 获取与输入按键序列相对应的候选字符串的方法与设备

Publications (1)

Publication Number Publication Date
WO2012139475A1 true WO2012139475A1 (zh) 2012-10-18

Family

ID=44570211

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/073589 WO2012139475A1 (zh) 2011-04-11 2012-04-06 获取与输入按键序列相对应的候选字符串的方法与设备

Country Status (2)

Country Link
CN (1) CN102184028A (zh)
WO (1) WO2012139475A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106527754A (zh) * 2016-10-31 2017-03-22 努比亚技术有限公司 一种优先字词排列方法、装置和一种移动终端

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102184028A (zh) * 2011-04-11 2011-09-14 百度在线网络技术（北京）有限公司 获取与输入按键序列相对应的候选字符串的方法与设备
CN102819567B (zh) * 2012-07-17 2016-04-20 重庆优腾信息技术有限公司 一种电子设备的联系人信息检索方法和系统
CN102902368A (zh) * 2012-10-12 2013-01-30 上海量明科技发展有限公司 一种输入法工具的实现方法及客户端
WO2014101224A1 (zh) * 2012-12-31 2014-07-03 Zhou Bin 数字终端上的字符输入方法、系统与数字终端
CN103699233B (zh) * 2013-12-20 2019-04-09 百度在线网络技术（北京）有限公司 字符串输入方法和输入装置
CN107340880B (zh) * 2016-05-03 2021-11-02 北京搜狗科技发展有限公司 一种联想输入方法、装置及用于实现联想输入的电子设备
CN107145242B (zh) * 2017-03-24 2020-09-29 北京百度网讯科技有限公司 纠错词展示方法、装置、设备及存储介质
CN107748784B (zh) * 2017-10-26 2021-05-25 江苏赛睿信息科技股份有限公司 一种通过自然语言实现结构化数据搜索的方法
CN110263337B (zh) * 2019-06-17 2023-06-20 北京百度网讯科技有限公司 输入行为识别方法、装置、设备和介质
CN111782060A (zh) * 2020-06-18 2020-10-16 维沃移动通信有限公司 对象显示方法、装置和电子设备
CN112684910A (zh) * 2020-12-29 2021-04-20 维沃移动通信有限公司 输入法候选词显示方法、装置及电子设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1153092A (ja) * 1997-07-31 1999-02-26 Nec Eng Ltd キー入力装置及びその入力方法並びにその入力方法をコンピュータに実行させるためのプログラムを記録した記録媒体
CN1655107A (zh) * 1999-05-27 2005-08-17 美国在线服务公司 自动更正键盘系统
CN1928860A (zh) * 2005-09-05 2007-03-14 日电（中国）有限公司 用于校正按键错误的方法、搜索引擎和搜索系统
CN101266520A (zh) * 2008-04-18 2008-09-17 黄晓凤 一种可实现灵活键盘布局的系统
CN102184028A (zh) * 2011-04-11 2011-09-14 百度在线网络技术（北京）有限公司 获取与输入按键序列相对应的候选字符串的方法与设备

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1282900A (zh) * 2000-09-07 2001-02-07 马晓光 一种汉字输入的智能化人机交互方法及键盘布局
CN101739143B (zh) * 2009-12-03 2014-05-07 深圳市世纪光速信息技术有限公司 文字输入方法及系统
CN101788879A (zh) * 2010-01-18 2010-07-28 孙国立 软键盘布局及扫描输入方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH1153092A (ja) * 1997-07-31 1999-02-26 Nec Eng Ltd キー入力装置及びその入力方法並びにその入力方法をコンピュータに実行させるためのプログラムを記録した記録媒体
CN1655107A (zh) * 1999-05-27 2005-08-17 美国在线服务公司 自动更正键盘系统
CN1928860A (zh) * 2005-09-05 2007-03-14 日电（中国）有限公司 用于校正按键错误的方法、搜索引擎和搜索系统
CN101266520A (zh) * 2008-04-18 2008-09-17 黄晓凤 一种可实现灵活键盘布局的系统
CN102184028A (zh) * 2011-04-11 2011-09-14 百度在线网络技术（北京）有限公司 获取与输入按键序列相对应的候选字符串的方法与设备

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106527754A (zh) * 2016-10-31 2017-03-22 努比亚技术有限公司 一种优先字词排列方法、装置和一种移动终端

Also Published As

Publication number Publication date
CN102184028A (zh) 2011-09-14

Similar Documents

Publication Publication Date Title
WO2012139475A1 (zh) 获取与输入按键序列相对应的候选字符串的方法与设备
KR100891358B1 (ko) 사용자의 다음 문자열 입력을 예측하는 글자 입력 시스템및 그 글자 입력 방법
JP4130846B2 (ja) 操作支援装置および操作支援方法
KR101878488B1 (ko) 대화 연관 컨텐츠 제공 방법 및 장치
WO2011150730A1 (zh) 一种用于英文与另一种文字混合输入的方法和设备
WO2016008452A1 (zh) 高效输入的预测方法和装置
WO2016178337A1 (ja) 情報処理装置、情報処理方法及びコンピュータプログラム
US11573989B2 (en) Corpus specific generative query completion assistant
WO2020026366A1 (ja) 特許評価判定方法、特許評価判定装置、および特許評価判定プログラム
CN103365573A (zh) 一种对多键输入字符进行识别的方法和装置
WO2022135474A1 (zh) 信息推荐方法、装置及电子设备
CN107748784A (zh) 一种通过自然语言实现结构化数据搜索的方法
CN107155121B (zh) 语音控制文本的显示方法及装置
CN103051969A (zh) 一种节目搜索方法及具有节目搜索功能的电视
JP2003030207A (ja) モバイル機用取扱問題解決処理システム及びモバイル機用取扱問題解決処理プログラム並びにデータベース検索プログラム並びに当該プログラムを記録したコンピュータ読み取り可能な記録媒体
CN104345896B (zh) 拼音文字词组输入方法及系统
TWI352970B (en) Voice input system and voice input method
WO2016131425A1 (zh) 滑行输入方法及装置
US8782067B2 (en) Searching method, searching device and recording medium recording a computer program
CN107797676B (zh) 一种单字输入方法及装置
JP5640700B2 (ja) 情報検索装置、検索入力方法、及び検索入力プログラム
JP5544003B2 (ja) 情報検索装置、情報検索システム、及び情報検索方法
CN103616962B (zh) 一种信息处理方法和装置
JP2020021455A (ja) 特許評価判定方法、特許評価判定装置、および特許評価判定プログラム
KR101355284B1 (ko) 터치스크린 환경에서 단어 추천 및 문장 완성 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12771547

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12771547

Country of ref document: EP

Kind code of ref document: A1