JP2000010973A - Character input system/method and recording medium which computer can read - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a Japanese input prediction system which can effectively use a limited prediction candidate selection area and can reduce manipulated variable on the candidate selection of a user by adopting a candidate display method and a selection method, which correspond to the character string length of a prediction candidate. SOLUTION: A character input system is provided with a retrieval means 4 retrieving a character string which matches an inputted read character string or whose front part matches from an input prediction dictionary 5 as a prediction character string, a display means 3 displaying the retrieved prediction character string, and selection means 2 and 6 selecting the range of the partial character string used for the prediction character string.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character processing system and method, and more particularly to a character input system and method for predicting a character string to be input from a part of a character string already input.

[0002]

2. Description of the Related Art In recent years, networks have been rapidly developed,
Communication using computers is becoming popular. Along with this, in addition to the traditional main uses of document creation, such as creating business documents using a desktop computer and writing a dissertation, more personal use, such as using e-mail using a portable information device and creating memos, has been made. In many situations, the opportunity to input Japanese sentences with a computer is increasing, and even people who are not skilled in keyboard operation are required to have a technique for accurately inputting information with few key touches.

On the other hand, various electronic devices such as computers have been reduced in size, and portable devices have been widely used. In a portable device, since the size of the input operation unit is small, the burden of inputting a document is further increased, and it is desired to reduce the burden associated with user input.

As one solution to such a request, a part of a character string to be input is inputted, and a character string table registered with priorities in advance or a past input character string stored with frequency is stored. Among them, there has been proposed a method of presenting a predicted character string in the order of priority and frequency (for example, Japanese Patent Application Laid-Open No. 60-105029).
JP, JP-A-61-74062, JP-A-61-74062
156468 and JP-A-2-131623).

In these methods, since there are a plurality of character strings to be predicted, the user needs to perform an operation of selecting a desired character string from these candidates. Conventionally, for this selection operation, a method of displaying a list of candidates and selecting a user from the list using a pointing device, a cursor key, or the like has been generally used in the same manner as the homophone selection operation in kana-kanji conversion. Was.

However, particularly in portable devices, the area for displaying these selection screens is limited, and the number and length of candidates to be displayed are limited. However, compared to homophone candidates in kana-kanji conversion, a character string generated by prediction has a larger number of candidates and a longer length. Therefore, in order to display these candidates in a limited screen, the number of predicted candidates must be narrowed down considerably, but advanced language analysis technology is required for high-precision prioritization Therefore, in reality, the user cannot present the desired result, that is, the recall rate is reduced.

Further, in the conventional method, there is only a method of employing the entire character string. For example, if you want to enter the string "Because a meeting will be held,"
And select "Candidates will be held."
Is displayed, first select this, then delete the part of "so," by the editing means, and enter "kara" again, and the procedure is complicated. .

[0008]

SUMMARY OF THE INVENTION In view of this point, the present invention has a prediction candidate selecting means which enables a user to select a desired character string with a small number of operations even if the selection area of the prediction candidates is small. It is an object of the present invention to provide a document input prediction system.

[0009]

According to a first aspect of the present invention, there is provided a retrieval means for retrieving a character string which matches or prefixes with an input reading character string from an input prediction dictionary as a prediction character string, and A character input system comprising: display means for displaying a string; and selection means for selecting a range of a partial character string to be used for the predicted character string.

A second aspect of the present invention is the character input system according to the first aspect, further comprising selecting means for selecting a character in the predicted character string and a character string ahead of the character.

According to a third aspect of the present invention, there is provided a display means for displaying character string division information for dividing the predicted character string into partial character strings together with the predicted character string, and selecting the divided area and the divided area ahead of the divided area. The character input system according to the first aspect of the present invention, further comprising:

According to a fourth aspect of the present invention, there is provided a prioritizing means for arranging the predicted character strings retrieved by the retrieval means in order of priority, and as a result of the prioritization, the number of candidates for the predicted character string is one. And selecting means for selecting the predicted character string without inquiring the user when the certainty factor of the predicted character string is larger than a predetermined threshold value. A character input system according to the invention.

According to a fifth aspect of the present invention, there is provided first display means for displaying the predicted character string for each partial character string; and first selecting means for selecting the partial character string and a partial character string ahead thereof. First predictive candidate selecting means comprising: a hierarchical layering means for layering the predicted character string; second displaying means for displaying the hierarchical layered character string at each stage; Second selection means for selecting a coded character string; and designation means for selectively designating the first or second prediction candidate selection means. A character input system according to a first aspect of the invention.

According to a sixth aspect of the present invention, there is provided first display means for displaying the predicted character string for each partial character string; and first selecting means for selecting this partial character string and a partial character string ahead thereof. First predictive candidate selecting means comprising: a hierarchical layering means for layering the predicted character string; second displaying means for displaying the hierarchical layered character string at each stage; Second selection means for selecting a coded character string; and designation means for automatically designating the first or second prediction candidate selection means. A character input system according to a first aspect of the invention.

According to a seventh aspect of the present invention, a character string that matches or starts with the input read character string is searched from the input prediction dictionary as a predicted character string, and the searched predicted character string is displayed. This is a character input method characterized by selecting a range of a partial character string to be used for.

According to an eighth aspect of the present invention, there is provided a retrieval procedure for retrieving a character string that matches or prefixes with an input reading character string from an input prediction dictionary as a prediction character string, and a display procedure for displaying the retrieved predicted character string. And a selecting step of selecting a range of a partial character string to be used for the predicted character string. A computer-readable recording medium storing a character input program to be executed by a computer.

According to the present invention, particularly when displaying and selecting a predicted character string having a long character string, there is provided a selection means for selecting a partial character string in the predicted character string. Even if the results are unintended,
If a character string intended for input is included in the predicted character string, it is possible to take out only that part and adopt it, reducing the load on input operations that had to be dealt with by redoing input etc. Can be.

In addition, as described above, by providing the second prediction candidate selecting means in addition to the first prediction candidate selecting means, especially in the case of a predicted character string having a long character string, the first prediction candidate selecting means can be used. In the case of a predicted character string having a short character string, the prediction candidate can be selected by the second prediction candidate selection means. The first prediction candidate selection unit can display a long prediction character string, particularly when the number of prediction candidates is small, thereby making full use of the limited selection area and obtaining a prediction result in a long unit. . On the other hand, when the number of prediction candidates is large, the second prediction candidate selection means can display a large number of prediction candidates without leaving a limited selection area. The probability that the predicted result is included is increased. Therefore,
By switching and using the first means and the second means based on the control of the input predicting means or the direct instruction of the user, it is possible to select candidates suitable for each characteristic. Thus, the number of operations performed by the user for input can be reduced.

In particular, when only one prediction result with extremely high certainty is obtained by the prioritizing means,
Since there is a means for adopting the prediction candidate as a selection result without inquiring the user, the trouble of selecting the candidate can be saved.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the character prediction input system of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a character prediction input system according to an embodiment of the present invention. The input unit 1 as an input means includes, for example, a keyboard, a mouse, a pen input device, etc., and inputs a character string or moves a cursor.
A command input such as an editing instruction such as insertion / deletion of a character string, an operation instruction such as display of a next candidate, selection of a candidate, etc. is taken in. After receiving and analyzing the information captured by the input unit 1, the control unit 2 sends necessary information to each processing unit. The processing result of each processing unit is returned to the control unit 2 again, and necessary information, for example, an input character string (hereinafter, referred to as an input character string) is output via the output unit 3 and displayed on a display or the like. Is done.

The input character string of the information input from the input unit 1 is sent to the input prediction unit 4 via the control unit 2.
The input prediction unit 4 generates a prediction candidate while referring to the input prediction dictionary 5, and returns a result to the control unit 2.

The input prediction dictionary 5 stores reading information, kanji kana notation information, part of speech information, notation frequency information, and reading frequency information. For example, for the notation “kana-kanji conversion”, the reading “kanakanjihenkan” and the part of speech “sa-hen” are given, and the numerical value representing the notation frequency is “60”, indicating the reading frequency. "12980, 890, 230, 90, 80, 8"
0, 80, 80 ". Here, the notation frequency information" 60 "is a numerical value indicating the frequency at which the notation" kana-kanji conversion "is used. It can be represented by the number of occurrences of the notation “Kana-Kanji conversion”. In addition, the reading frequency information is delimited by “,”, and in this example, eight numerical values for eight partial character strings are given to nine character readings “Kanakanjihenkan”. The first “12980” is the frequency of the partial reading character string “kana”, the second “890” is the reading frequency of “kana”, and thereafter “230” is “kanakan” and “90”. “Kana Kanji”, “80” “Kana Kanji”, the next “8
“0” is a numerical value indicating the frequency of reading “Kanakanjikan”, the next “80” is “Kanakanjikan”, and the next “80” is a numerical value indicating the frequency of reading “Kanakanjikan”. Can be represented by, for example, the number of occurrences of a character string such as “Kana”, “Kanaka”, or “Kanakan” in a corpus to which a reading for a certain corpus is given.

Next, the control unit 2 sends the generated prediction candidates to the prediction candidate selection unit 6. The prediction candidate selection unit 6 generates a display image of each candidate by using the long unit prediction candidate selection unit 7 or the short unit prediction candidate selection unit 8 according to an instruction of the candidate selection method instruction unit 9 described later, To the control unit 2 in the form of Note that, when information indicating a request request for a candidate selection method is input from the input unit 1,
It is driven by the control unit 2 and provides an interface for instructing a candidate selection method described later.

Further, the control section 2 displays on the output section 3 a candidate selection button (for example, "Hold" in the area 203 in FIG. 2), which will be described later, received from the prediction candidate selection section 6, and the input section 1
Wait for information to obtain a later-described selection range from. This information is sent to the prediction candidate selection unit 6, and when the prediction candidate selection unit 6 determines that the candidate has been determined, the information is received and a determined character string, which will be described later, is displayed in the character input area of the output unit 3.

The above is a brief description of each component of the character prediction input system of the present invention, and an overview of the processing flow of these components. Through these processes, an interface screen as shown in FIG. 2 is provided by the output unit 3 to the user of the system.

The interface screen is divided into three areas, for example, a command area 201, a character input area 202, and a candidate selection area 203. In the command area 201, menus such as "File", "Edit", "Settings", and "Help" are prepared. Instructions relating to document management, instructions relating to editing operations, character input methods, candidate selection methods, and the like are respectively provided. Instructions for various settings, guidance on operation methods, and the like can be provided. Character input area 20
Reference numeral 2 denotes an area for displaying a currently-read character string ("key") or a character string determined in the past ("" or the like). The candidate selection area 203 is an area for displaying and selecting a candidate for a character string mixed with kanji or kana (“such as a meeting will be held,” etc.) predicted from the currently input reading character string.

The control unit 2 provides the above interface screen by controlling the input unit 1, the output unit 3, the input prediction unit 4, the prediction candidate selection unit 6, and the candidate selection method instruction unit 9.

Then, next, regarding the operation of the control unit 2,
This will be described in more detail with reference to the flowchart of FIG. Note that, for simplification of the description, the character string input by the user is only a reading character string. It is assumed that the input prediction unit 4 generates, as a predicted character string, a character string mixed with kanji or kana, which has a character string that exactly matches or has a forward match with the input read character string.

First, input information input from the input unit 1 is analyzed (S301). As input information, a character string to be read is input to the character input area 202 by a soft keyboard or handwritten character recognition (S
302: Yes), otherwise (N of S302)
o), that is, a case where a command is input to the command area 201.

Further, when a command is input in the command area 201 (No in S302),
When there is a setting request of the candidate selection method from the “setting” menu of the command area 201 (Yes in step S303)
The processing is performed separately for s) and other cases (No in S303).

In other cases (No in S303),
For example, other processing (S304) is performed using the "edit" menu in the command area 201, for example, when there is an editing request such as deletion, copy, and paste of a character string. still,
Since this other processing (S304) has no direct relation to the present invention, a detailed description is omitted.

When the input information is a reading character string (S3
(02, Yes) performs input prediction, prediction candidate selection, and the like by a series of processes from S306 to S310. A series of these processes will be described later.

If the input information is a request for setting a candidate selection method (Yes in S303), the candidate selection method designating section 9 is driven in S305 to set the candidate selection method.
In S305, first, a window for setting a method of selecting a prediction candidate is displayed on the output unit 3 as shown in FIG. The user selects one of “automatic (0)”, “long unit display (1)”, and “short unit display (2)”
Press the "OK" button. Here, "Length display" means
The display of prediction candidates with the number of characters as long as possible, such as phrase units or sentence units, and "short unit display" indicate that the prediction candidates are displayed with the number of characters reduced, such as word units. It is for giving instructions. Further, “automatic” means that the character input system of the present invention determines whether to adopt “long unit display” or “short unit display”. The initial value is set to, for example, “automatic”.

When the user presses the "OK" button in FIG. 4, the setting values relating to the method of selecting a prediction candidate are stored in the setting storage memory managed by the prediction candidate selection unit 6. As shown in FIG. 5, the setting storage memory is stored in the order of attribute name-attribute value.
Stored as attribute value of attribute name “external setting value”. As attribute values of the attribute name “external setting value”, “0”, “1”,
Takes any one of "2" and corresponds to "automatic", "long unit display", and "short unit display", respectively. The attribute name "internal set value" is set by the input prediction unit 4, and this attribute value is either "1" or "2", and is "long unit display" or "short unit display", respectively. Corresponding to

In the case of this embodiment, since "automatic (0)" is selected in FIG. 4, the external setting value in FIG. 5 is "0".
become. Also, here, the internal set value, “Length unit display (1)” indicates a case where it becomes 1. How to determine the internal setting value will be described later.

When the set values are stored, the candidate selecting method instructing section 9 closes the window shown in FIG. 4 and terminates a series of processes related to candidate selection shown in S305. Next, a series of processes of S306 to S310, whose description has been suspended, will be described.

First, in S306, the control unit 2 acquires a read character string from the input unit 1 and stores it in an initialized read character string buffer (not shown). The read character string is displayed at the cursor position in the character input area 202, and the cursor is advanced. Here, the cursor refers to “=” next to “key” in the character input area 202 of FIG. FIG. 2 shows that the read character string is an undetermined character string.
An underline as shown in the example of “Kaigi” in the character input area 202 is added.

Subsequently, the control unit 2 transfers the control to the input prediction unit 4 and performs input prediction. In response, the input prediction unit 4
Makes an input prediction based on the read character string stored in the read character string buffer (S306). The input prediction unit 4
A search is made from the input prediction dictionary 5 for a predicted character string that matches or precedes the read character string stored in the buffer. The search result is prioritized using probability information or the like, and a prediction candidate is generated. The prioritization using the probability information and the like means the priority ordering of the retrieved kanji kana written character string using the frequency information for the read character string and the read frequency information for at least a part of the partial character string of the read character string. (Japanese Patent Application No. Hei 10-0389)
90 publication). The input prediction unit 4 outputs “L” when the length L _y of the received character string is shorter than a predetermined threshold value L _min or when no search result is obtained from the input prediction dictionary 5. A result indicating "no prediction candidate" is returned to the control unit 2. In S308, the presence or absence of the prediction candidate is determined, and if there is no prediction candidate, the process returns to S306 again to acquire the next read character string. In this way, the input of the reading character string continues until the prediction candidate is obtained.

For example, _assuming that L _min = 3, no prediction candidate is obtained at the stage of inputting “KA” and “I” for the input “KAIGI” illustrated in FIG. Next, "gi"
Is input, L _y ≧ L _min is satisfied, and
A prediction candidate having a character string of a dictionary heading that completely matches or prefixes with the reading character string “Kaigi” is searched, and by prioritizing the search results, a prediction result as shown in FIG. 6 is obtained.
It is stored in a prediction candidate buffer (not shown) and passed to the prediction candidate selection unit 6.

In FIG. 6, the candidate numbers indicate the priority of the obtained prediction candidates. The certainty factor (CF) is a numerical value of 0 <CF ≦ 1, which is calculated based on the probability information described in the input prediction dictionary 5, and indicates the reliability of each candidate. The higher the confidence, the higher the reliability of the candidate. The predicted character string is separated by a special delimiter ("/") for each word constituting the predicted character string. Furthermore, each word is a character string ("conference")
) And the part of speech name of the word (“[name]” etc.). For example, the first candidate in FIG. 6 indicates a character string “Because a meeting is held,” and this character string is a noun “meeting”, a particle “wo”, a verb stem “holding”, and a ending “shi”. , The auxiliary verb stem "ma", the ending "su", the particle "no", and the reading point ",".

The input prediction unit 4 compares the number of generated candidates with the candidate number upper limit C _max determined by the size of the selection candidate area 203 and the character string length of each candidate. If the number of generated candidates is smaller than _Cmax, it is determined to be “display in units of length”, and the value of the attribute name “internal set value” of the setting storage memory shown in FIG. 5 is set to “1”. If not,
"2" indicating "short unit display" is stored.

When the prediction candidate is obtained as described above, the process proceeds to S309, in which the prediction candidate selection unit 6 is driven to prompt the user to select a prediction candidate. Details of the operation of the prediction candidate selection unit 6 will be described later with reference to the flowchart of FIG. Next, when the prediction candidate selection is completed and the control unit 2 receives the determined character string from the prediction candidate selection unit 6, the control unit 2 proceeds to S31.
With 0, a confirmed character string is displayed in the character input area 202. For example, in the above example, if the character string "Because a meeting will be held" is adopted, the underlined "Kaigi" part shown in FIG. Is overwritten, and the cursor is advanced by the number of characters. Note that the undetermined character string may have already been deleted, as in the short unit display described later. In this case, the determined character string is displayed at the cursor position, and the cursor is simply advanced by the number of characters. When the process of FIG. 3 is completed, the process enters a standby state of waiting for a new input by the user.
1 is performed.

Next, the processing of the prediction candidate selection (S309) performed by the prediction candidate selection unit 6, whose description has been suspended, will be described in detail with reference to the flowchart of FIG. The prediction candidate selection unit 6 first receives a prediction candidate as shown in FIG. 6 from the control unit 2 (S701).

Next, a prediction candidate selection method is determined (S70).
2). In S702, first, the number of candidate prediction candidates is checked, and if the number of candidates is 1 and there is a candidate whose certainty factor is larger than a threshold CF _min predetermined in the initial setting of the system, the determination is made. It is determined to be possible (S703).
The threshold value CF _min is usually set to a value very close to 1, for example, when a fixed phrase or the like is predicted, a certainty factor exceeding this may be obtained. When it is determined that the determination is possible, the process proceeds to S704, where the prediction candidate selection unit 6 concatenates the character strings constituting the candidates, returns this as a confirmed character string to the control unit 2, and ends the processing. As a result, the prediction result can be adopted without making the user troublesome in selecting candidates.

If the determination is not possible, one of "long unit display (1)" and "short unit display (2)" is determined with reference to the setting storage memory shown in FIG. First, the attribute value of the attribute name “external setting value” is checked, and the attribute value is “1”,
If it is any of “2”, it is determined as “long unit display (1)” and “short unit display (2)”, respectively. If the attribute value is “automatic (0)”, the attribute value of the attribute name “internal setting value” is further checked. If “1”, “long unit display”; Is determined.

If "Length unit display" is determined (S70)
5) In step S706, the long unit prediction candidate selection unit 7 is driven.
To S708 are performed. When the “short unit display” is determined, the short unit prediction candidate selection unit 8 is driven, and the processing of S709 to S713 is performed.

First, S7 performed by the long unit prediction candidate selection unit 7
The processing from 06 to S708 will be described. The output unit 3 is driven via the prediction candidate selection unit 6 and the control unit 2 to display a long unit in the candidate selection area 203 (S706). In long unit display, the layout of candidate arrangements is determined based on the shape and size of the candidate selection area, the character string length of each candidate, and the priority. At this time, if all the candidates cannot be displayed, the number of candidates may be reduced based on the priority and certainty of the candidates. When the layout is determined, each candidate is converted into a button-shaped display image. For this purpose, the received predicted character string of each prediction candidate is divided into words based on word delimiters, and the part of speech names as shown in FIG. 6 are removed. Then, with respect to the constituent character strings of each extracted word, word delimiters (for example, vertical bar "|")
Creates a display image by performing string concatenation while inserting.

As a result, a candidate selection button as shown in the candidate display area 203 of FIG. 2 is displayed. In addition,
Here, the divided area is created in units of words, but it is of course possible to create the divided area in units of phrases or in units of characters. In the case of a phrase unit, one independent word + 0 or more attached word chains may be set as a division unit with reference to the part of speech information. In the case of character units, it is not necessary to insert a delimiter. FIGS. 8A and 8B show examples of a phrase-based candidate selection button and a character-based candidate selection button when no delimiter is displayed.

Next, the input unit 1 waits for an input from the input unit 1 via the prediction candidate selection unit 6 and the control unit 2.
When an area surrounding the pressed position of the above candidate selection button, for example, “S” is acquired, “Conference is held” is acquired as a selection range (S707).

Next, a determined character string is determined from the pressed position (S708). First, it is possible to determine which candidate has been selected from the pressed position. Next, it can be determined which of the candidate divided regions has been selected. Then, a fixed character string is obtained by concatenating the partial character string of the divided area and the partial character string displayed in the divided area ahead thereof from the head. For example,
In the example of FIG. 2, the first prediction candidate “meeting /
When the sixth divided area “su” of “shi / ma / su / so /,” is pressed, “meeting”, “wo”, “holding”, “shi”,
The character strings of “ma” and “su” are sequentially concatenated to obtain a confirmed character string “meeting / a / held / held / ma / su”. As a result,
Assuming that the user is expecting a meeting to be held, the user can use a single button operation to select a meeting to be held. The desired result can be obtained.

On the other hand, in the conventional method, there is only a method of adopting the entire character string. Therefore, either give up the selection of candidates and continue inputting, or select "I will hold a conference," and then edit Procedures such as deleting the part of "so" by means and inputting "kara" again are necessary, and the operation is complicated.

When the candidate selection button is divided and displayed in units of phrases as shown in FIG. 8A, a confirmed character string can be obtained in the same manner as described above. FIG.
Even in the case where character units are displayed without delimiters as in (b), it is possible to determine which character has been selected from the pressed position, so that the character and the character string preceding it are connected from the beginning. By doing so, a confirmed character string can be obtained.

With the above processing, the determined character string in the “long unit display” is determined, and the control returns to the prediction candidate selection unit 6. Then, S704 described above is executed, and the process of selecting a prediction candidate (S309) ends.

Next, the prediction candidate selecting section 6 sets the "short unit display"
The operation of the short unit prediction candidate selection unit 8 that is driven when it is determined (No in S705) will be described. The short unit is a word unit or a phrase unit. In this case, the method can be implemented by a method substantially similar to the method disclosed in “Method and Apparatus for Displaying Conversion Candidates of Reading Input Character String” in Japanese Patent Application Laid-Open No. 9-319738.

First, the short unit prediction candidate selection unit 8 converts the received prediction character string of each prediction candidate into a hierarchical data structure (hereinafter, referred to as a hierarchical structure) (S70).
9). For example, the prediction candidates in FIG. 6 are converted into the hierarchical structure shown in FIG.

Next, the short unit prediction candidate selection unit 8 drives the output unit 3 via the prediction candidate selection unit 6 and the control unit 2 to display the short unit in the candidate selection area 203 (S71).
0). In the short unit display, the candidates are sequentially displayed stepwise while referring to the hierarchical structure. In the first hierarchy, that is, in the short unit display, a node connected by a direct arc (a portion indicated by an arrow) from the start node is selected as a candidate. In the example of FIG.
And "Skeptics" are selected. Then, based on the shape and size of the candidate selection area, the character string length of each candidate, and the priority,
The layout of the candidate arrangement is determined. When the layout is determined, each candidate is converted into a button-shaped display image. At this time, if an arc extending from each candidate is connected to a node that is not the end node, a symbol (for example, “$”) indicating that there is a subsequent candidate is added after the character string to be displayed. As a result, a selection candidate button as shown in FIG. Here, since the prediction candidate “meeting” has a subsequent prediction result,
Although the symbol "@" is added, "skeptic" has no subsequent prediction result, and thus only the character string is displayed.

Next, the short unit prediction candidate selection unit 7 waits for an input from the input unit 1 via the prediction candidate selection unit 6 and the control unit 2, and displays the selection candidate on the candidate selection area 203 by The system obtains a selection candidate by a pressing operation by an input or the like (S711).

Further, the presence or absence of a succeeding candidate is checked (S712). If the selected candidate does not have a subsequent prediction candidate, this is determined as a confirmed character string, and control returns to the prediction candidate selection unit 6 (No in S712). Then, S704 described above is executed to select a prediction candidate (S3
09) is ended. For example, this corresponds to the case where the candidate “skepticism” is selected in FIG.

On the other hand, when the selected candidate has a succeeding prediction candidate, the selected candidate is referred to as a partially determined character string, and the output unit 3 is driven via the prediction candidate selection unit 6 and the control unit 2. Then, the partial confirmed character string is displayed in the character input area 202, and the cursor is advanced by the number of characters (S713). At this time, the undetermined character string is entered in the character input area (202 in FIG. 2).
If it is in the state indicated by, it is deleted. Next, the process returns to S710 to display and select a subsequent prediction candidate, and the above processing is repeated until there is no subsequent prediction candidate. For example, in FIG. 10A, when the candidate “meeting” is selected, the unconfirmed character string “Kaigi” remains in the input character area 202, and this is deleted. Overwrite "Meeting". And S71
Returning to 0, the short unit display of the second stage of the hierarchical structure is performed. The second row is displayed as shown in FIG. It should be noted that a special candidate "CAN" is displayed in the second and subsequent stages.
"CEL" is prepared. This is used when there is no candidate expected by the user among the candidates of the row. If the candidate “CANCEL” is selected, the process goes to S71.
In 2, the empty character string is determined as the confirmed character string,
It is determined that there is no subsequent candidate, and the control returns to the prediction candidate selection unit 6 (No in S712). Then, S704 described above is performed.
Is performed, and the process of selecting a prediction candidate (S309) ends. With the special candidate “CANCEL”, it is possible to avoid losing the previous selection result. For example, FIG.
If “CANCEL” is selected in the state of 0 (b), “Conference” which is the result of the selection before that is already adopted and displayed in the character input area 202.
Subsequently, the user can input the next reading character string.

The above is the processing for determining the character string in the short unit prediction candidate selection unit 8. When the character string determination process is completed under the above-described conditions, the control returns to the prediction candidate selection unit 6, and S704 is executed to select a prediction candidate (S309).
Is completed.

In the above description, for convenience of explanation, the character string serving as the base of input prediction has been limited to a hiragana character string, that is, only a read character string. However, the present embodiment can be naturally expanded even when a character string serving as a base for input prediction is a character string mixed with Chinese characters.

In fact, in a system having a character input mechanism employing the online handwritten character recognition technology, an incomplete kanji-kana mixed character string is converted into a complete kanji-kana mixed character using an input prediction dictionary 5 as shown in FIG. It has a mechanism to convert to a column. For example, it is troublesome to input "meeting" by hand, so that "meeting" is obtained from input such as "meeting". Essentially, such conversion is realized by expanding the dictionary headings to character strings mixed with kanji and kana, instead of being limited to kana character strings.

Of course, in order to reduce the storage capacity, various compression methods have been proposed instead of the simple table format as shown in FIG. 11, but logically, the table of FIG. 11 is expressed. No change.

Therefore, also in the present embodiment, if a character string mixed with kanji and kana is recognized in the reading information of the input prediction dictionary 5, it is possible to extend the character string serving as the base of the input prediction to a character string mixed with kanji and kana. it can. In this case, the constraint condition imposed on the character string of the input prediction candidate is that when a part of the character string is returned to the kana, the character string completely matches or prefixes with the character string mixed with the kanji and kana characters. The other processing portions are processes for the prediction result, and therefore do not need to be changed. Therefore, a case where a character string serving as a prediction base is expanded to a character string mixed with kanji and kana can also be said to be within the scope of the present embodiment.

Although the present invention has been described by exemplifying only Japanese input, the present invention is not limited to this. For example, FIG.
The input prediction dictionary 5 may have English or the like such as “meeting” in the notation section 1. The character input described in the above embodiment can be distributed as a program that can be executed by a computer, stored in a recording medium such as a floppy disk, a hard disk, a CD-ROM, or a semiconductor memory.

A program for executing the character input described in the above embodiment and an input prediction dictionary are stored in any one of the recording media. Then, according to this program, the above-described character input is executed by the CPU (arithmetic control device) with respect to the character string input from the input unit and the input prediction dictionary read from the recording medium. By doing so, the character input of the present invention can be performed using a normal personal computer.

[0067]

As described above, according to the present invention,
In particular, when displaying and selecting a prediction unit in a long unit, since a means for indicating a partial character string in the predicted character string is provided, even if the predicted character string as a whole is an unintended prediction result of the user, the character If a character string intended for input is included in the column, only that part can be extracted and adopted. As a result, it is possible to reduce the load of the input operation as compared with the conventional method in which the input method has to be redone.

Further, in addition to the above-described means for predicting a long unit, there is provided a prediction candidate selecting means for a short unit which employs the entire predicted character string as a selection result. By switching and using it based on control or a direct instruction of the user, it is possible to select a candidate suitable for each feature. In other words, when the number of prediction candidates is small, by displaying a long predicted character string, a limited selection area can be fully utilized and a long unit prediction result can be obtained. On the other hand, when the number of prediction candidates is large, a large number of prediction candidates can be displayed without leaving a limited selection area, so that the probability that the intended prediction result is included in the displayed candidates is high. Become. That is, as a result, the number of operations performed by the user for input can be reduced.

Furthermore, when only one prediction result with extremely high certainty is obtained, there is provided a means for adopting the prediction candidate as a selection result without inquiring the user. Can also.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of a Japanese input prediction system according to the present invention.

FIG. 2 is a diagram of an interface screen provided to a user.

FIG. 3 is a flowchart illustrating a processing operation of a control unit.

FIG. 4 is a window screen view for setting a method of selecting a prediction candidate.

FIG. 5 is a diagram showing description contents of a setting storage memory.

FIG. 6 is a diagram illustrating an output example of an input prediction unit.

FIG. 7 is a flowchart illustrating a processing operation of a prediction candidate selection unit.

FIG. 8 is a diagram illustrating an example of a candidate selection button in long unit display.

FIG. 9 is a diagram illustrating an example of a hierarchical structure.

FIG. 10 is a diagram showing an example of a candidate selection button in short unit display.

FIG. 11 is a diagram showing an example of an input prediction dictionary whose heading is a character string mixed with kanji and kana.

[Explanation of symbols]

 Reference Signs List 1 input unit 2 control unit 3 output unit 4 input prediction unit 5 input prediction dictionary 6 prediction candidate selection unit 7 long unit prediction candidate selection unit 8 short unit prediction candidate selection unit 9 candidate selection method instruction unit 201 command area 202 character input area 203 Candidate selection area

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Yoshimi Saito 1st Toshiba R & D Center, Komukai-ku, Kawasaki-shi, Kanagawa Prefecture (72) Inventor Hideki Hirakawa Toshiba Komukai-shi, Kawasaki-shi, Kanagawa 1-cho, Toshiba R & D Center F-term (reference) 5B009 MA04 MA11 ME17 5B075 ND03 PQ02 PQ36 PQ46 PQ74 QM03 QP02 UU05

Claims (8)

[Claims] A search unit that searches a character string that matches or prefixes with the input read character string as a predicted character string from an input prediction dictionary; a display unit that displays the searched predicted character string; A selection unit that selects a range of a partial character string to be used for the character string. 2. The character input system according to claim 1, further comprising: selecting means for selecting a character in the predicted character string and a character string ahead of the character. 3. Display means for displaying character string division information for dividing the predicted character string into partial character strings together with the predicted character string; and selecting means for selecting the divided area and the preceding divided area. The character input system according to claim 1, further comprising: 4. A prioritizing means for arranging the predicted character strings searched by the searching means in order of priority, and as a result of the prioritization, the number of candidates for the predicted character string is one; 2. The character according to claim 1, further comprising: selecting means for selecting the predicted character string without inquiring the user when the certainty factor of the predicted character string is larger than a predetermined threshold value. Input system. 5. A system comprising: first display means for displaying the predicted character string for each partial character string; and first selecting means for selecting the partial character string and a partial character string ahead of the partial character string. 1, a prediction candidate selecting unit, a hierarchizing unit that hierarchizes the predicted character string, a second display unit that displays the hierarchized hierarchized character string at each stage, Second selecting means for selecting,
2. The character input system according to claim 1, further comprising: a second prediction candidate selection unit including: a first prediction candidate selection unit; and a designation unit that selectively designates the first or second prediction candidate selection unit. 3. 6. A system comprising: first display means for displaying the predicted character string for each partial character string; and first selecting means for selecting the partial character string and a partial character string ahead of the partial character string. 1, a prediction candidate selecting unit, a hierarchizing unit that hierarchizes the predicted character string, a second display unit that displays the hierarchized hierarchized character string at each stage, Second selecting means for selecting,
2. The character input system according to claim 1, further comprising: a second prediction candidate selecting unit that includes: a designation unit that automatically designates the first or second prediction candidate selection unit. 3. 7. Searching a character string that matches or prefixes with an input read character string as a predicted character string from an input prediction dictionary, displays the searched predicted character string, and uses the predicted character string. A character input method, wherein a range of a partial character string to be selected is selected. 8. A search procedure for searching a character string that matches or prefixes with an input read character string as a predicted character string from an input prediction dictionary, a display procedure for displaying the searched predicted character string, A computer-readable recording medium storing a character input program to be executed by a computer, the method comprising: selecting a range of a partial character string to be used for the character string.