JP2001014309A - Device and method for converting character and recording medium recording program for character conversion - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily input the sentence of a response to the sentence of another person by inputting the sentence prepared by the other person, counting how many times a word appears in that sentence, finding the priority of conversion on the basis of the number of times of appearance, preparing a dictionary for response and performing character conversion while using this dictionary for response. SOLUTION: The sentence prepared by the other person is inputted, how many times a word appears in that sentence is counted, reading KANA is added, the priority of conversion is found on the basis of the number of times of the appearance, the dictionary for response is prepared and character conversion is performed while using the dictionary for response. In such a device, a word appearance count part 3 analyzes the syntax of the read sentence, decomposes it into words, counts the number of times of the appearance for the unit of a word and prepares an appearance times table for the unit of a word. A dictionary preparing part 4 for response adds reading KANA for the unit of a word on the basis of the appearance times table, generates character conversion support data for response after ranking based on the number of times of appearance of words, sorts the data in the order of reading KANA, number of times of appearance and character code and stores them in a dictionary 4-1 for response.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character conversion device and a method for inputting text in a word processor, e-mail, various editors and the like, and a recording medium on which a computer program for realizing the method is recorded. .

[0002]

2. Description of the Related Art Conventionally, when inputting a sentence using a word processor, e-mail, various editors, or the like, a reading kana is input from an input device such as a keyboard and character conversion is performed to convert the kana into appropriate characters. I was typing. If there is more than one candidate for the character corresponding to the reading kana, if the first candidate character is the character to be converted, it is sufficient to proceed to the input of the next character, but if the first candidate is not the character to be converted Is a general conversion method in which a conversion key is pressed several times until an appropriate character is displayed, or a list of conversion candidates is displayed and a character to be converted is selected from the list.

[0003]

As computers are becoming more sophisticated and lower in price as of today, not only offices,
With the rapid spread of computers at home, the improvement of the communication infrastructure of information networks, the increase in communication speed and the reduction of communication costs, etc. The number of people using word processors and e-mail is increasing dramatically.

[0004] For this reason, opportunities to create sentences using a keyboard or the like have been greatly increased. Along with that
Opportunities to create texts for responding to texts created by others, such as e-mail, are increasing. When inputting a sentence of a response to a sentence created by another person, the word used in the sentence is often used, but when inputting the word, the character conversion registered in its own conversion dictionary is used. In many cases, it is troublesome to press the conversion key many times or to display a list of conversion candidates and select an appropriate character from the list. for that reason,
A word that is often used in the source text to respond to,
There is a demand for a technical means that is preferentially converted as the first candidate and reduces the time and effort for character conversion.

[0005]

[MEANS FOR SOLVING THE PROBLEMS] When inputting a sentence for responding to a sentence created by another person, since the sentence is about the same topic as the sentence of the other person, the word used in the sentence is used. Is often used. However, when performing the character conversion by inputting the reading kana of the word used in the sentence, the first candidate of the conversion candidate character for the reading kana in the conversion dictionary of the respondent is used in the sentence. I don't know if it's a word. In particular, if the content of the sentence is related to a specific topic or technology, words that are not commonly used are often used, so the priority of the responder in the conversion dictionary is low. In many cases, it is troublesome to press the conversion key many times to search for a corresponding word, display a list of conversion candidates, and select from the list.

According to the present invention, in order to reduce the complexity of character conversion when inputting a response document, a sentence created by another person is input, the number of appearances of a word appearing in the sentence is calculated, and the reading kana is calculated. Add, determine the conversion priority based on the number of appearances,
Create a response dictionary. By performing character conversion using this response dictionary, words used in another person's text will be given priority in conversion based on the number of appearances. Is more likely to be the first candidate for conversion, thereby reducing the burden of character conversion.

A character conversion device according to the present invention is a character conversion device for inputting a reading kana and performing character conversion using a character conversion dictionary, wherein a sentence reading unit for reading a sentence, a word used in the sentence, A character conversion assisting unit for response that creates character conversion auxiliary data for a response that recognizes the reading kana for that and reflects it in the priority of character conversion candidates,
Character conversion of an input sentence related to the read sentence is performed using the response character conversion assisting unit.

With this configuration, the priority of character conversion is changed according to the number of appearances of a word used in a document from which a response is made, and the priority of a word used in the document increases. In the character conversion when inputting a response document, the probability that the character to be converted is the first candidate or a character corresponding thereto becomes high, so that the character conversion key is pressed many times, It is possible to considerably reduce the trouble of displaying a character candidate list and selecting from the list.

A character conversion device according to the present invention is used in a system for transmitting and receiving electronic texts via a network. In the character conversion device for inputting a reading kana and performing character conversion using a character conversion dictionary, a character conversion device is provided. A sentence reading unit for reading a sentence received in a word, a word used in the sentence, a word appearance count unit for counting the number of appearances in word units, and a reading kana added to the word unit, and the word unit Based on the number of appearances, the conversion priority is determined, a response dictionary creation unit that creates a response dictionary, a response dictionary that stores the response dictionary data, the character conversion dictionary, and the response dictionary. And a character conversion unit for performing character conversion based on the character conversion.

With this configuration, character conversion is performed in accordance with the number of times of use of words used in e-mail, conversation contents, and documents in a format such as a homepage exchanged via a network such as e-mail, electronic chat, and the Internet. Since the priorities are changed and the priorities of the words used in the document become higher, the time and labor required for the character conversion at the time of inputting the response document can be considerably reduced.

According to a character conversion method of the present invention, in a character conversion method of inputting a reading kana and performing character conversion using a character conversion dictionary, a step of reading a sentence; Creating a response character conversion auxiliary data for recognizing the reading kana and reflecting the same in the priority of the character conversion candidates; and using the response character conversion auxiliary unit to input the character of the input sentence related to the read sentence. The method includes a step of performing a conversion.

With this configuration, the priority of character conversion is changed according to the number of appearances of a word used in a response source document, and the priority of a word used in the document is increased. Because the probability that the character to be converted becomes the first candidate or a similar character is high in character conversion when inputting the response document, the conversion key must be pressed many times during character conversion. And the trouble of displaying a character candidate list and selecting from the list can be considerably reduced.

A computer-readable recording medium storing a program to be executed by a computer according to the present invention records a program for inputting a reading kana and causing a computer to perform a step of performing character conversion using a character conversion dictionary. A computer-readable recording medium, comprising: reading a sentence; and character conversion auxiliary data for a response that recognizes a word used in the sentence and a reading kana for the word and reflects the word in a priority order of a character conversion candidate. And a step of performing character conversion of an input sentence related to the read sentence using the response character conversion assisting unit.

With such a configuration, the priority of character conversion is changed according to the number of appearances of a word used in a document as a response source, and the priority of a word used in the document becomes higher. In the character conversion when inputting the response document, whether the character to be converted is the first candidate,
A character conversion device that can reduce the trouble of pressing the conversion key many times in character conversion or displaying a character candidate list and selecting from among them, because the probability of becoming similar is high. Can be realized.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (Embodiment 1) The principle and basic operation flow of the present invention will be described with reference to the basic configuration diagram of the present invention in FIG. 1 and a flowchart showing the basic processing flow in FIG. I do.

First, the text reading unit 1 reads a text from the text file 1 assigned in advance (step S1). Here, the sentence is assumed to be a digitized text file. However, the present invention is not limited to this, and the sentence file 1 may not be in a file format. In step S1, a specific document is input as having been assigned in advance, but if there are several documents, a step of designating a sentence may be provided before step S1.

Next, the response character conversion assisting unit 3X recognizes the words used in the read text and their reading kana, reflects them in the priority of character conversion candidates, and converts the character conversion auxiliary data for response. It is created (step S2).

Here, the reading kana means all characters and characters to be input when performing character conversion by a normal word processor or the like.
It includes numbers and symbols, and is not limited to hiragana. In addition, the priority order may be simply the number of appearances, or may be divided into several priorities from the number of appearances. Up to this point, the response dictionary creation process at the stage before input of the response text is performed.

FIG. 1 shows a word appearance counting section 3, a response dictionary creating section 4, and a response dictionary 4-1 as an example of a specific configuration of the response character conversion assisting section. The word appearance counting unit 3 performs syntax analysis on the read sentence to decompose the sentence, counts the number of appearances in word units, and creates an appearance number table in word units. In word decomposition, words whose endings change depending on how they are used, such as verbs, adjectives, adjective verbs, etc., are regarded as the same word even if they have different endings. The response dictionary creator 4 adds the reading kana for each word based on the appearance frequency table, assigns priorities based on the number of appearances of the word, generates response character conversion auxiliary data, and generates the generated data. Are sorted in the order of kana, the number of appearances, and the character code, and stored in the response dictionary 4-1. The addition of the reading kana requires reverse processing to the conversion of the reading kana into characters. In the configuration diagram, the addition of the reading kana is performed by converting a word and a reading kana conversion table (not shown) inside the response dictionary creating unit.
However, it may be provided outside the response dictionary creation unit 4 as a reading kana conversion dictionary, or the character conversion dictionary 5-1 may be provided only with an index capable of performing character conversion from the reading kana. Alternatively, an index that can be converted from a word to a reading kana may be added, and the reading kana may be obtained from the word using the character conversion dictionary 5-1. In this configuration diagram, the configuration diagram in which the response dictionary is created from the response character conversion auxiliary data is shown. However, the configuration is not particularly limited to the configuration in which the response dictionary is created. You can have it in memory as a table and use it,
The contents of the response character conversion auxiliary data are directly reflected in the contents of the character conversion dictionary.
Only -1 may be used.

Next, creation of a response document is started. First, a reading kana of a word to be input as a sentence is input using the input device 6 such as a keyboard (step S3). The input device 6 is not limited to a keyboard, but may be anything that can input characters. The input kana is subjected to character conversion using the original character conversion dictionary 5-1 and the newly created response dictionary 4-1 and the result is output (step S4). When converting characters,
The conversion by the response dictionary 4-1 is prioritized, and the response dictionary 4-
If not, the conversion is performed using the original character conversion dictionary 5-1. In the present embodiment, the conversion result is displayed on the display device 7 such as a display device, and is simultaneously output to the file as the created document 5-2. However, the present invention is not limited to this. You may pass it. While the creation of the response sentence is not completed (step S
5: NO) returns to step S4 to continue character input,
The creation of the response sentence ends (step S5: YES).
Then, the response sentence creation processing ends.

The created response dictionary 4-1 is a dictionary for responding to an input sentence. If it is used temporarily to respond to the input sentence, it is automatically deleted after the response sentence is created. You may do so. If you want to save the response dictionary for reuse, automatically generate and save the response dictionary name based on the file name of the input document, or enter the response dictionary name and save it You may make it. In the former case, if the name given to the response dictionary is called a tag, the file name of the input document becomes the tag. Also,
The response dictionary creation process and the response document creation process do not necessarily need to be performed consecutively, but may be performed individually.

If there are a plurality of texts on the same theme or from the same person as the original document of the response dictionary 4-1 created first, and it is desired to create a response dictionary based on all the texts, The information of the document input thereafter may be stored in the response dictionary created in step S2, or a plurality of sentences may be input simultaneously to create the response dictionary. Target documents may be grouped and input from a plurality of sentences according to some conditions. In this case, the configuration is as shown in FIG.

Prior to the processing by the text reading unit 2, the response dictionary creation condition input unit 8 inputs response dictionary creation conditions from an input device 6 such as a keyboard, and stores it in the response dictionary creation condition storage unit 2-1. Store. The text reading unit 2 receives the response dictionary storage condition from the response dictionary creation condition storage unit 2-1 and reads only the target text file from the text file 1 according to the specified condition. The following processing is the same as the configuration diagram of FIG.

FIG. 4A shows an example of designation of a response dictionary creation condition when the input document is a text file. The OBJECTS control statement used here specifies a file to be input. The specified file name is not limited to one, and a plurality of file names may be listed. When the same character is included in the file name, the file name may be specified by a wild card as in (2).

The response dictionary creation condition does not necessarily need to be determined by the text reading unit 2. When the response dictionary creation process and the character conversion process are performed separately, the response dictionary 4-1 is prepared for each document in advance. In accordance with the response dictionary storage conditions, the corresponding response dictionaries are combined into one to create a combined response dictionary 4-2 as shown in FIG. You may.

When creating a response sentence, the response dictionary 4
-1 or the original dictionary is used, but the way of reflecting the character information determined at the time of conversion may be applied only to the response dictionary, only the original character conversion dictionary, or both. May be reflected in the dictionary.

As a result, the priority of character conversion is changed according to the number of appearances of a word used in a document from which a response is made, and the priority of a word used in the document becomes higher. In character conversion when inputting a document, the character to be converted becomes the first candidate or the probability that it is ranked according to it increases, so you need to press the conversion key many times in character conversion, It is possible to considerably reduce the trouble of displaying the list and selecting from the list.

(Second Embodiment) In the first embodiment, a case where a document is already digitized and stored as a file in a storage device has been described. However, a sentence to be input is not necessarily digitized. However, the input document may be a paper-based document such as a book or a magazine, may be a voice, or may be an Internet HTML document file read via a line.

In this embodiment, a case where the input document is a paper-based document will be described with reference to FIG. The only difference from FIG. 1 is that of the document input processing. Therefore, the same parts as in FIG. 1 are assigned the same reference numerals and explanations thereof will be omitted as appropriate.

The document reading unit 2 is provided with an OCR (Optical Charac
ter Reader) functions and devices. First, the document reading unit 2 inputs a character image of the paper-based document 9 using the OCR function, and performs a character recognition process 2a by off-line processing using the character recognition dictionary 2-2.
The document 9 is extracted as a digitized text document. For the parts of the document 9 such as figures other than the characters, the processing is skipped and the processing is not performed. The character recognition dictionary 2-2 means a dictionary necessary for the character recognition processing 2a, and the type of the character recognition dictionary may differ depending on the character recognition processing method. In addition, when a tag name is assigned to the response dictionary, since the input is a paper-based document and does not have a file name serving as a tag name, it is necessary to provide a tag name input unit and input.

Thus, even if the input document is a paper-based document such as a book or a magazine, it is possible to create a response dictionary and reduce the trouble of character conversion.

(Embodiment 3) Next, the case where the input document is a voice will be described with reference to FIG. The only difference from FIG. 1 is that of the document input processing. Therefore, the same parts as in FIG. 1 are assigned the same reference numerals and explanations thereof will be omitted as appropriate.

It is assumed that the document reading section 2 has a voice recognition processing 2b function. First, the text reading unit 2 inputs voice using a voice input device 10 such as a microphone,
A voice recognition process 2b is performed using the voice recognition dictionary 2-3,
Convert audio to digitized text documents. The voice recognition process includes the process of digitizing analog voice data input from the microphone. The voice recognition method can recognize and parse the voice using a known voice recognition method and convert it to a text document. Is not particularly limited as long as it is appropriate. The speech recognition dictionary 2-3 includes all dictionaries necessary for speech recognition, such as a speech unit dictionary and a speech pattern dictionary required for speech recognition. In addition, when a tag name is given to the response dictionary, since the input does not have a file name which becomes a tag name by voice, it is necessary to provide a tag name input means and input.

Thus, the input document is, for example,
It is possible to create a response dictionary and reduce the time required for character conversion, even when directly inputting by voice or by inputting by playing a tape recording the speech data of the lecture .

(Embodiment 4) Next, a case where an Internet HTML document file in which an input document is read via a line is read by a browser will be described with reference to FIG. The only difference from FIG. 1 is the document input processing part, and therefore, the same parts are assigned the same reference numerals, and description thereof will be omitted as appropriate.

HTML document reading unit 2c of browser 13
Reads the HTML document file 1-1 on the web server 11 via the line 12 based on the specified URL, analyzes the read HTML document for the control statement portion of the HTML document, and includes the HTML document in the HTML document. The text and graphics are displayed on a display device 7 such as a display in accordance with a designated HTML control statement. At the same time, the text document extracting unit 2e- incorporated in the HTML document reading unit 2c
1 extracts only the text part from the HTML document excluding the control sentence and the figure, and passes it to the document reading unit 2.

When inputting an HTML document, a response document to the HTML document is not always created on the browser.
The response dictionary 4-1 is created and saved, and after browsing the document with the browser, the response dictionary 4-1 is specified by an editor such as a word processor or an e-mail to create a document. Is also good.

In FIG. 8, the processing after the document reading unit 2 is performed outside the browser 13. However, when a response document is created on the browser 13, the response document may be incorporated in the browser 13.

As described in the first embodiment, when the response dictionary creation condition processing is added, the OBJECT control statement to be specified is as shown in (3) of FIG. Of course, a plurality of addresses may be listed. The response dictionary 4-1
, A URL may be used as a tag name, or a title name usually added to an HTML document may be used as a tag name. However, if the URL is too long,
Since the length of the file name of the response dictionary 4-1 is limited, it may be necessary to shorten the file name.

Thus, even when the input document is an HTML document on the Internet, it is possible to create a response dictionary and reduce the trouble of character conversion.

(Embodiment 5) Next, an embodiment of a character conversion apparatus used in a system for transmitting and receiving an electronic document via a network will be described with reference to the configuration diagram of FIG. 9 taking an electronic mail as an example. The same parts as those in FIG. 8 are denoted by the same reference numerals, and the description of the overlapping parts will be appropriately omitted.

In the present embodiment, an example in which the present embodiment is combined with software called a mailer for operating an electronic mail will be described.

First, the user logs in to the mailer 15 to activate the mailer 15. When the mailer 15 is activated, the mail transmission / reception unit 2d of the mailer 15 sends an e-mail addressed to the logged-in person in the e-mail 1-2 stored in the e-mail server 14 via the line 12 based on the logged-in ID. And displays a list of e-mail addressed to the person on a display device 7 such as a display. Select the e-mail you want to see from the e-mail list and display its contents. If the reply mode is specified while referring to the e-mail, the mailer 15 starts the editing editor and performs a response dictionary creation process. First, when the editing editor is activated, the mode is changed to a reply e-mail input mode. The functions described so far are functions included in a normal mailer, and are not particularly illustrated in the configuration diagram.

The character conversion device is incorporated in the e-mail editor, and when entering a mode for replying to the original e-mail, passes the e-mail to the text reading section 2 as an input document. The response dictionary creation processing as described in the first embodiment is performed, and the reply electronic mail 5-3 is executed.
Create When the reply e-mail has been edited and the mailer's reply button is pressed, the reply e-mail is returned.

Thus, when creating a reply to an e-mail, a response dictionary for the original e-mail is automatically created, and the priority of conversion of words frequently used in the original e-mail is increased. Character conversion can be made faster and easier.

Although the processing for one e-mail has been described here, it is not necessary to create a response dictionary for only one e-mail. For example, one response dictionary 4-1 may be created from a plurality of e-mails by specifying the response dictionary creation conditions described in the first embodiment.

FIG. 4B shows an example in which a response dictionary creation condition is specified by electronic mail. The e-mail is added with information such as a sender ID, a recipient ID of the destination, and a title of the e-mail called a subject, and the response dictionary creation condition is designated using the information. For example, as shown in (1), a document to be input may be specified by a specific sender, and all the e-mails of the person may be grouped and input, or as shown in (2), the group name may be specified in advance. If a group name is specified, all the e-mails belonging to the group may be grouped and input. Alternatively, as shown in (3), one or more e-mail titles may be specified. Alternatively, if it is desired to group by topic, as in (4), all e-mails that include a specific term (here, “business”) in the subject as the title may be grouped and input. Good. As described above, the input means for grouping the e-mails is provided, the grouping conditions can be set, and the response dictionary 4-1 is created by automatically grouping the e-mails according to the set conditions. Good. An electronic bulletin board is similar to an electronic mail except that the article information is not addressed to a specific individual. In this way, the response dictionary 4-1 created using the subject name, the sender ID, and the like as the tag name according to the additional information of the electronic mail.
It is possible to create

Also, the response dictionary creation processing and the character input / conversion processing can be performed separately, and response dictionaries are created and stored in advance according to topics and technologies, and an appropriate response dictionary is designated at the time of character input. By doing so, an arbitrary response dictionary can be used even in cases other than when creating a reply e-mail. If it is possible to switch the response dictionary even during the creation of the e-mail, it is possible to always select an appropriate response dictionary even when topics change in the middle of the e-mail.
It is possible to reduce the trouble of character conversion.

(Embodiment 6) Next, an embodiment of a character conversion apparatus used in a system for transmitting and receiving an electronic document via a network will be described with reference to FIG. 10 using an electronic chat (hereinafter referred to as chat) as an example. explain. 8, the same parts as those in FIG. 8 have the same numbers, and the description of the overlapping parts will be omitted as appropriate.

In the present embodiment, the present invention is applied to a character conversion device used in a chat system, and a case where the input text is a message of each user displayed on the screen when using the chat will be described. I do.

First, the user logs in to the chat to activate the chat. When the chat starts, the user specifies the name of a channel to which the user wants to participate in the chat, joins the channel, and starts a conversation. In the chat, a plurality of users can participate in one channel, and a chat message (hereinafter, referred to as a message) inputted by all the participating users is transmitted to IRC (Internet R).
eal time Chat) Displayed on the screen of each user in real time along with the user name and time via the server.

At the time of joining the channel, the input message has not yet been displayed on the screen, and the message starts from the time of joining. Therefore, it is not possible to immediately create a response dictionary. Even if a response dictionary is created immediately after starting a chat, it is not very effective if there are few messages to be input. Also, in the case of chat, messages are exchanged in real time, so if the topic changes on the way, the response dictionary created before that may not work effectively.

In order to solve these problems, it is desirable that the creation of the response dictionary is not limited to one point in time, but is variable depending on conditions. Therefore, in order to make the creation and updating of the response dictionary variable according to various conditions, it is preferable that the timing of creating the response dictionary and the target and range of the input document can be set in advance. FIG.
FIG. 5 is a configuration diagram when a response sentence is created with a response dictionary creation condition and character conversion is performed in electronic chat. The start processing from logging in, specifying a channel name, and starting chat is an electronic chat function, and is not shown here.

First, at the time when the user logs in to the chat and joins the channel, no message has been input yet, and a response dictionary cannot be created. Therefore, when inputting one's own message, character conversion is performed using only the original dictionary 5-1. When someone enters a message, the entered message is
It is transmitted to the C server 16 and added with the login ID of the sender and the input time, and is stored as the chat message 1-3. Come. The chat message transmitting / receiving unit 2e displays the received message on the display device 7 such as a display, and at the same time, passes the message to the text reading unit 2.

The text reading unit 2 accumulates the message in the chat message accumulation file 2-4, and if the response dictionary creation condition is specified in the response dictionary creation condition storage unit 2-1, it is specified. When the condition is satisfied, the process proceeds to a response dictionary creation process. If the response dictionary creation condition is to automatically create a response dictionary unconditionally each time a message is received, the response dictionary creation condition input unit 8 and the response dictionary creation condition storage unit 2-1 And the chat message storage file 2-3 can be omitted. Here, the input of the response dictionary creation condition is performed before the chat is started. However, the input may be performed at the start of the chat or may be performed during the creation of the response document.

Since the chat message is input in real time, when the message arrives at the chat message transmitting / receiving section 2e, the control is automatically transferred to the chat message transmitting / receiving section 2e even in the character conversion processing mode. When the chat message transmitting / receiving unit 2e receives the message, the message is passed to the text reading unit 2, and when the response dictionary creation condition is satisfied, a response dictionary creation process is executed, and when the response dictionary creation condition is not satisfied. Transfers control to the character conversion process without creating / updating a response dictionary.

The following are some examples of setting response dictionary creation conditions.

FIGS. 4C to 4C show examples of setting of a response dictionary creation condition when the present embodiment is applied to an electronic chat.

In the example of (1), in a chat, since a message comes in real time, a response dictionary is created /
This is an example of response dictionary creation conditions for designating that the update timing is real-time with a TIMING control statement and for creating and updating a response dictionary each time a message is received. As a result, the latest response dictionary is always created, so that character conversion can always be performed according to all message contents.

In the example of (2), by specifying the sender ID in the OBJCTS control statement, only the message of a specific sender is input to create a response dictionary, and the response dictionary is created.
In this example, the update timing is set in units of 10 minutes. In a chat, since a sender ID is added to a message, only a message with a specific sender ID can be used as an input sentence. The time specified by INTERVAL is checked by providing the text reading unit 2 with a time monitoring unit (not shown). The time monitoring unit has a clock unit and a response document creation time storage unit. The time monitoring unit may start the response dictionary creation process every predetermined time,
Each time a message is received, it may be determined whether 10 minutes have elapsed and the processing may be started. By the time monitoring unit,
When it is determined that the time has elapsed, the text reading unit 2
Retrieves only the specified message from the chat message storage file 2-4 and starts a response dictionary creation process. If 10 minutes have not elapsed, the text reading unit stores the received message in the chat message storage file 2-4.
And the control is shifted to the character conversion process without performing the response dictionary creation process. The message stored in the chat message storage file may be only the message of the designated sender. As a result, it is possible to limit the input message for creating the response dictionary to the message of the specific sender, which is the main part of the conversation. In addition, when the response dictionary is updated in real time, the response dictionary is frequently updated although the contents of the response dictionary do not change much for each message. Therefore, it is possible to prevent the response dictionary from being updated.

In the example of (3), by specifying a group name in the OBJECTS control statement, only messages from senders in a specific group are input for creating a response dictionary, and the timing of creating and updating the response dictionary is determined by the message. Is an example in which the storage of data increases by 5 kilobytes. The text reading unit 2 or the response dictionary creation condition storage unit 2-1 has a group table (not shown) in which the correspondence between the group name and the sender ID is defined. A message to be an input for creating a response dictionary is identified from the group table, and the response dictionary is created and updated every time the message increases by 5 kilobytes. To check the capacity increase of the message of the ADD control statement, the text reading unit 2 is provided with a capacity increase monitoring unit (not shown).
Each time a message is received, the number of bytes that have been received since the last response dictionary was created is counted,
Start processing when the specified amount is exceeded. The capacity monitoring unit has capacity measuring means and means for storing up to which message processing has been performed. As a result,
The RVAL designation has a problem that the response dictionary is updated even if there is no received message for a predetermined time. However, the response dictionary can be updated every time a predetermined amount of messages increases by the ADD designation. The meaningful update of the response dictionary can be prevented.

The example of (4) is an example in which, for all received messages, the latest 20 kilobyte message is input in real time to create a response dictionary. CAPA
Checking the capacity increase of the CITY control statement message
The sentence reading unit 2 is provided with a capacity calculation monitoring unit (not shown). Each time a message is received, the received message is stored in the chat message storage file 2-4, and the message is taken out from the file in the order of the newest one. The number is calculated, and up to the point where the total capacity exceeds the specified capacity is defined as the input text for creating and updating the response dictionary. The capacity calculation unit has a capacity measurement unit, and the text reading unit 2 has a function of reading messages from the chat message storage file 2-4 in the latest message order. In the case where the story is changed for a long time or in the middle of a story, even if a response dictionary including old messages is created, many words that are no longer used may be included. Since only a message having a predetermined capacity can be input to the response dictionary, character conversion can always be performed according to the latest topic.

The tag name of the response dictionary is preferably a channel name, but may be a sender ID or a group name specified by an OBJECTS control statement.

Thus, when inputting a message in the electronic chat, an appropriate response dictionary can always be created and updated by inputting a predetermined message at a predetermined timing, so that appropriate character conversion is always performed. Can create a response dictionary.

(Embodiment 7) Next, a recording medium in which a program for realizing a character conversion device according to an embodiment of the present invention is described, as shown in the example of the recording medium shown in FIG.
Not only the portable recording medium 19 such as the ROM 19-1 and the floppy disk 19-2, but also any other storage device 18 stored at the end of a communication line, or the recording medium 21 such as a hard disk or a RAM of a computer. When using the data display device according to the present invention, the computer 2
0 and is executed on the main memory.

[0066]

According to the character conversion device of the present invention, when creating a response document for any text or message, the text or message is input and a response dictionary is created, Because the transliteration of words used in sentences or messages takes precedence,
The possibility that converted character candidates will be displayed as first candidates based on the input kana is increased, and complicated operations such as repeatedly pressing a character conversion key and selecting one from a character candidate group are greatly reduced. Can be done.

[Brief description of the drawings]

FIG. 1 is a basic configuration diagram of the present invention.

FIG. 2 is a flowchart showing a basic processing flow of the present invention.

FIG. 3 is a configuration diagram of a character conversion device according to a first embodiment of the present invention (with a response dictionary creation condition);

FIG. 4 is an example of a response dictionary creation condition according to the present invention.

FIG. 5 is a configuration diagram of a character conversion device according to the first embodiment of the present invention (dictionary merging based on response dictionary creation conditions).

FIG. 6 is a configuration diagram of a character conversion device according to a second embodiment of the present invention.

FIG. 7 is a configuration diagram of a character conversion device according to a third embodiment of the present invention.

FIG. 8 is a configuration diagram of a character conversion device according to a fourth embodiment of the present invention.

FIG. 9 is a configuration diagram of a character conversion device according to a fifth embodiment of the present invention.

FIG. 10 is a configuration diagram of a character conversion device according to a sixth embodiment of the present invention.

FIG. 11 is an exemplary diagram of a storage medium.

[Explanation of symbols]

1 Sentence File 1-1 HTML Sentence File 2 Sentence Reading Unit 2a Character Recognition Processing 2b Speech Recognition Processing 2c HTML Sentence Reading Unit 2c-1 Text Sentence Extraction Unit 2d Mail Transmission / Reception Unit 2e Chat Message Transmission / Reception Unit 2-1 Response Dictionary Creation Conditions Storage Unit 2-2 Character Recognition Dictionary 2-3 Speech Recognition Dictionary 3X Response Character Conversion Aid Unit 3 Word Appearance Counting Unit 4 Response Dictionary Creation Unit 4-1 Response Dictionary 4-2 Merged Response Dictionary 5 Character Conversion Unit 5-1 Dictionary 5-2 Creating sentences 6 Input device such as keyboard 7 Display device such as display 8 Response dictionary creating condition input section 9 Paper-based sentences 10 Voice input device such as microphone 11 Web server 12 Line 13 Browser 14 Mail Server 15 Mailer 16 IRC server 17 Electronic chat 18 Store at the end of communication line It was another storage device 19 CD-ROM or floppy disk, etc. of a portable storage medium 19-1 CD-ROM 19-2 floppy disk 20 computer 21 RAM / hard disk storage medium on the computer

Claims (5)

[Claims] 1. A character conversion apparatus for inputting a reading kana and performing character conversion using a character conversion dictionary, comprising: a sentence reading section for reading a sentence; recognizing a word used in the sentence and a reading kana for the word; A response character conversion assisting unit for creating a character conversion assisting data for a response to be reflected in the priority order of the character conversion candidates, and using the response character conversion assisting unit, an input sentence associated with the read sentence. A character conversion device for performing character conversion of: 2. A character reading device used in a system for transmitting and receiving electronic texts via a network, for inputting a reading kana and performing character conversion using a character conversion dictionary, and for reading texts received via the network. A word used in the sentence, a word appearance count unit that counts the number of appearances per word, and a reading kana added to the word unit, based on the number of appearances of the word unit, conversion priority A response dictionary creating unit that determines the order and creates a response dictionary; a response dictionary that stores the response dictionary data; and a character conversion that performs character conversion based on the character conversion dictionary and the response dictionary. A character conversion device comprising a part. 3. A response dictionary creation condition input unit for inputting conditions for creating the response dictionary, and a response dictionary creation condition storage unit for storing the response dictionary creation conditions, wherein the response dictionary creation 3. The character conversion device according to claim 1, wherein a response dictionary is created or updated based on a condition. 4. A character conversion method for inputting a reading kana and performing character conversion using a character conversion dictionary, comprising the steps of: reading a sentence; recognizing a word used in the sentence and a reading kana corresponding thereto; Creating a character conversion auxiliary data for a response to be reflected in the priority of the conversion candidate, and performing character conversion of an input sentence related to the read text by using the character conversion auxiliary unit for a response. Character conversion method to use. 5. A computer-readable recording medium in which a program for causing a computer to execute a step of inputting a reading kana and performing a character conversion using a character conversion dictionary is recorded, wherein: a step of reading a sentence; Recognizing the word used and the reading kana for the word, and creating character conversion auxiliary data for a response to be reflected in the priority of the character conversion candidate, and reading the character conversion auxiliary data using the response character conversion auxiliary unit. And performing a character conversion of an input sentence related to the sentence.