JP2004240859A - Paraphrasing system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To realize a system which intelligibly paraphrases a phrase likely to require specialized knowledge. <P>SOLUTION: A dictionary database 31 is provided which associates a headword with a paraphrased word representing the headword by another expression, and indicates the words. Furthermore, a user information storage table 32 is provided which indicates whether or not a headword is to be paraphrased to a paraphrased word for each user. If text is inputted together with user information from an input part 11, a conversion part 22 refers to the dictionary database 31, and searches a headword matching an arbitrary phrase in the text. Furthermore, if a matching headword is found, the conversion part 22 refers to the user information storage table 32 to examine whether the user requires the conversion of the headword, and if required, converts the arbitrary phrase into paraphrased words. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a paraphrase system for translating text into paraphrases suitable for a user, and in particular, to understand foreign words and technical terms according to the level of the user (proficiency, possessed knowledge, physical limitations, etc.). It relates to a paraphrasing system that is easily paraphrased.
[0002]
[Prior art]
Conventionally, there has been a technique for switching a display screen according to a user's age, visual acuity, and proficiency (for example, see Patent Document 1).
In this document, the user sets the age so that the kanji to be used is restricted for elementary school children, and the display screen is changed, such as changing the size of characters to be used by performing eyesight measurement for elderly people. Performing techniques are disclosed.
[0003]
[Patent Document 1]
JP 2000-305746 A
[0004]
[Problems to be solved by the invention]
However, in the above-described conventional technology, the displayed content is uniform according to the level set by the system, and cannot be customized in detail according to the user's preference and proficiency.
[0005]
In recent years, in the field of computers and the like, foreign words (Katakana) and words that require specialized knowledge have been frequently used, and especially in the elderly, etc., there is a problem that understanding of sentences is hindered. is there. In such a case, a mechanism for paraphrasing difficult-to-understand expressions is necessary, but such a paraphrasing system has not been realized.
[0006]
[Means for Solving the Problems]
The present invention employs the following configuration to solve the above-described problem.
<Configuration 1>
A dictionary database that associates headwords with paraphrases that express the headwords in different expressions, and for any words and phrases, refers to the dictionary database, and if there is a matching headword, paraphrases the words and phrases A conversion system, comprising: a conversion unit that converts a word into a word and outputs the word.
[0007]
<Configuration 2>
In the paraphrase system according to Configuration 1, a user information storage table indicating whether or not a headword is paraphrased for each user, and, when a user is designated, an arbitrary information based on user information corresponding to the user. A paraphrase system comprising a conversion unit for converting a phrase into a paraphrase.
[0008]
<Configuration 3>
3. The paraphrasing system according to Configuration 2, further comprising a paraphrasing information setting unit that, when the user specifies whether to paraphrase the headword, reflects the specified content in the user information storage table.
[0009]
<Configuration 4>
In the paraphrasing system according to the configuration 2 or 3, when a word in the document provided by the user matches a headword in the dictionary database, a proficiency learning unit that sets the headword to be not paraphrased in the user information storage table. A paraphrasing system comprising:
[0010]
<Configuration 5>
In the paraphrase system according to the configuration 2 or 3, if the word in the document provided by the user matches the paraphrase in the dictionary database, the proficiency learning unit sets the headword as paraphrase required in the user information storage table. A paraphrasing system comprising:
[0011]
<Configuration 6>
In the paraphrase system according to any one of the constitutions 1 to 5, an input / output device connected via a communication line to the dictionary database and the conversion unit is provided, and the input / output device converts an arbitrary phrase together with a paraphrase request into the conversion unit. , And configured to receive the paraphrase result from the conversion unit.
[0012]
<Configuration 7>
In the paraphrase system according to any one of the constitutions 1 to 6, the dictionary database including the paraphrase corresponding to the condition based on the context relation of the headword, and the context relation of the phrase to an arbitrary phrase are shown. Based on the information, the dictionary database is referred to, and if there is a headword whose information satisfies the condition based on the contextual relationship, a conversion unit is provided that converts the headword into a paraphrase corresponding to the headword. A paraphrase system to do.
[0013]
<Configuration 8>
The paraphrasing system according to any one of Configurations 1 to 7, further comprising a conversion unit configured to morphologically analyze the input character string to extract words, and to convert a headword into a paraphrase based on the extracted words. Paraphrase system characterized by the following.
[0014]
<Configuration 9>
In the paraphrasing system according to any one of the constitutions 1 to 7, the character string is increased one character at a time from an arbitrary character in the document, and the character string is compared with a headword. A paraphrase system comprising a conversion unit that determines that the phrase is a phrase.
[0015]
<Configuration 10>
10. The paraphrasing system according to any one of Configurations 1 to 9, wherein a portion converted into a paraphrase is displayed differently from a portion not converted.
[0016]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail using specific examples.
<< Specific Example 1 >>
<Constitution>
FIG. 1 is a configuration diagram showing a specific example 1 of the paraphrasing system of the present invention.
The illustrated system is configured by a computer, and includes an input / output device 10, a processing device 20, and a storage device 30. The input / output device 10 is a functional unit that inputs a document to be paraphrased and outputs a paraphrase result, and includes an input unit 11 and an output unit 12. The input unit 11 has a function of performing a text document input process such as a text input process using a keyboard or a pointing device such as a mouse, a text input process using a scanner and a character recognition processing unit, and a text input process using a microphone and a voice recognition process. ing. The input unit 11 has a function of inputting paraphrase information individually set by the user. The output unit 12 has a function of performing display on a display device, conversion into audio, audio output, and output to a file, and outputs a processing result of text input from the input unit 11.
[0017]
The processing device 20 includes an arithmetic device, a memory, a control unit, and the like, and has a function of executing a process of setting a user's proficiency level and rephrasing a text input from the input unit 11 into another expression. are doing.
The processing device 20 includes a paraphrase information setting unit 21 and a conversion unit 22. The paraphrase information setting unit 21 is a functional unit that sets the proficiency of each user input from the input unit 11 and stores paraphrase information of a specific word in a user information storage table 32 described later.
[0018]
The conversion unit 22 refers to a later-described dictionary database 31 and a user information storage table 32 in the storage device 30 for an arbitrary word in the text input from the input unit 11, and determines that there is a matching headword. Has a function of converting the phrase into a paraphrase based on the user information in the user information storage table 32, and includes a morphological analysis unit 221, a dictionary search unit 222, and a text conversion unit 223.
[0019]
The morphological analysis unit 221 is a processing unit that performs a known morphological analysis for performing a process of dividing a text into words. The dictionary search unit 222 is a functional unit that performs processing for acquiring information associated with a headword with reference to the dictionary database 31 for the words divided by the morphological analysis unit 221. The text conversion unit 223 is a functional unit that performs a process of paraphrasing text based on the dictionary search result of the dictionary search unit 222.
[0020]
The storage device 30 includes a storage device such as a hard disk device, an optical disk device, or a semiconductor memory, and includes a dictionary database 31 and a user information storage table 32. The dictionary database 31 is a database that stores paraphrase data. The user information storage table 32 is a table for storing user proficiency level information.
[0021]
FIG. 2 is an explanatory diagram of the dictionary database 31 and the user information storage table 32.
As shown in the figure, the dictionary database 31 includes a headword storage unit 311, a paraphrase storage unit 312, and a condition storage unit 313, and one record is formed by the information of the user information storage table 32 corresponding to these fields. . The headword storage unit 311 stores words (headwords) that require paraphrase. Further, the paraphrase storage unit 312 stores a phrase (paraphrase) in which each headword is paraphrased with another expression. Further, the condition storage unit 313 is information for setting a condition for applying the paraphrase processing, and in this specific example, indicates the difficulty of the headword. Here, the difficulty level indicates the degree of specialization of the word. In general, the higher the difficulty, the harder it is to understand, and a paraphrase is necessary. In this specific example, it is assumed that the proficiency level and the difficulty level correspond to each other. Further, the user information storage table 32 indicates, for each piece of identification information (ID) of each user, information on whether or not to paraphrase each headword.
[0022]
<motion>
FIG. 3 is a flowchart showing the operation of the first embodiment.
First, a user using the present apparatus performs a user authentication process by inputting a user ID or the like from the input unit 11 (step S101). Then, selection is made as to whether to set paraphrase information or to perform paraphrase processing (step S102). In step S102, in the case of "Y", the operation of the conversion unit 22 is performed, and in the case of "N", the operation of the paraphrase information setting unit 21 is performed.
[0023]
1) When setting paraphrase information (in the case of "N" in step S102)
The setting of the paraphrase information includes a method of designating the overall level according to the criteria of the device, and a method of setting whether or not it is necessary to paraphrase each word.
[0024]
1-1) First, a method of setting paraphrase information by specifying the overall level will be described. Here, it is assumed that level 1 is defined as beginner, level 2 as intermediate, and level 3 as advanced.
[0025]
When the user selects and inputs his / her own level through the input / output device 10, the paraphrase information setting unit 21 sets this information in the user information storage table 32. For example, when the user selects level 2 (step S103), words having a difficulty level of level 2 or higher, that is, words of level 2 and level 3 are to be paraphrased, and the paraphrase information setting unit 21 sets the headwords of these levels. Is added to the field of the user information storage table 32 corresponding to the information (step S104). In FIG. 2, the user information storage table 32 of ID1 shows a state where the user of ID1 has selected “intermediate” = level 2.
[0026]
1-2) Next, a method of setting paraphrase information for each word will be described.
There are cases where it is desired or not desired to paraphrase a specific word regardless of the level of the user. In such a case, a word and information on whether to paraphrase the word are input through the input / output device 10 (step S103). Thereby, the paraphrase information setting unit 21 gives information on whether or not to paraphrase the word to the user information storage table 32 of the specified word (step S104). In FIG. 2, the user information storage table 32 of ID2 is a word of level 2 after the user of ID2 selects “intermediate” = level 2 and paraphrases the word of “level 1” “choriso”. This shows a state in which “Confid” is set to be unnecessary.
[0027]
2) When performing paraphrasing processing (in the case of “Y” in step S102)
When performing the paraphrasing process, the user inputs a text to be paraphrased from the input unit 11 (step S105). Thereby, the paraphrasing process is performed on the text in the conversion unit 22 (step S106), and the processing result is output from the output unit 12 by means such as a screen display (step S107). Thereafter, when processing a newer text (step S108), the process returns to step S105. Otherwise, the user logs out (step S109) and ends the process.
[0028]
Next, the operation of the paraphrasing process in step S106 will be described in detail.
FIG. 4 is a flowchart illustrating the operation of the paraphrasing process.
First, the morphological analysis unit 221 performs a morphological analysis on the input text (step S111). Next, the dictionary search unit 222 extracts one word from the morphological analysis result by the morphological analysis unit 221 (step S112), and searches the dictionary database 31 using the word as a key (step S113). If the word is not registered in the dictionary database 31, the word is stored in a buffer (not shown) in the conversion unit 22 (steps S114 and S115). If the word is registered in the dictionary database 31 as a headword, the corresponding paraphrase is stored in the buffer according to the information in the user information storage table 32 of the user (step S116). Next, it is determined whether the word extracted in step S112 is the last word of the text (step S117). If the word is not the last word, the process returns to step S112, and the above-described steps S112 to S115 (or step S116). Is repeated.
[0029]
FIG. 5 is a diagram illustrating a restaurant menu as an example of a document.
Hereinafter, the flow of the process when the user of ID2 performs the paraphrase process using this text will be specifically described.
[0030]
FIG. 6 is an explanatory diagram of the result of morphological analysis of one sentence of the text in FIG.
FIG. 7 is an explanatory diagram of the contents of the work buffer.
FIG. 8 is an explanatory diagram of the menu after the paraphrasing process.
[0031]
When performing the paraphrasing process, first, "chorizo" is read from the morphological analysis result (TX61) in FIG. Next, as shown in FIG. 2, when the dictionary database 31 is searched, it can be seen that the field of "chorizo" of ID2 is set to "o", that is, that paraphrase is required. Therefore, "dry sausage" which is a paraphrase of "chorizo" is stored in the buffer. That is, TX71 “dry sausage” in FIG. 7 is stored in the buffer.
[0032]
Since the user of ID2 has the proficiency level of 2, the word "chorizo" is a word of a difficulty level lower than the proficiency level of the user, but the paraphrase is performed by giving priority to the information in the user information storage table 32. Will be.
[0033]
Next, “enter” is read from the morphological analysis result shown in FIG. 6, and the dictionary database 31 is similarly searched. Since "enter" is not registered in the dictionary database 31, "enter" is registered in the buffer as it is (the state of TX72 in FIG. 7). Then, the same processing is performed on “Fu beans” to “Confi”.
[0034]
TX73 in FIG. 7 indicates the contents of the buffer at the time when the paraphrasing process for the sentence TX61 in FIG. 6 is completed. That is, paraphrase processing is performed on “lagou” and “poiblond” in the TX 61, and “confid” is output as it is because the user with ID 2 does not need to paraphrase. FIG. 8 shows the result of performing such a paraphrasing process on all the sentences in the menu shown in FIG. When the paraphrasing process is completed for all the sentences of the input text, the text conversion unit 223 sends the processing result to the output unit 12, and the output unit 12 displays the processing result on a screen or the like.
[0035]
<effect>
As described above, according to the specific example 1, the dictionary database 31 indicating the paraphrase for the headword is provided, and the vocabulary in the input sentence is converted into the paraphrase based on the dictionary database 31. Expressions that are difficult to understand, such as technical terms, can be paraphrased into expressions that are easy to understand according to the user's proficiency.
[0036]
Further, since the user individually sets whether or not to paraphrase each word, fine customization can be performed according to the user's preference and proficiency.
[0037]
<< Specific Example 2 >>
In the specific example 2, the user inputs a document indicating the proficiency level, and based on the document, sets whether or not the paraphrase is necessary in the user information storage table.
[0038]
<Constitution>
FIG. 9 is a configuration diagram of the specific example 2.
The illustrated system includes an input / output device 10, a processing device 20a, and a storage device 30. Here, the configurations of the input / output device 10 and the storage device 30 are the same as those in the first embodiment, and thus description thereof will be omitted. The processing device 20a includes a conversion unit 22 and a proficiency learning unit 23. That is, in the specific example 2, a skill learning unit 23 is provided in place of the paraphrase information setting unit 21 in the specific example 1. Here, since the conversion unit 22 is the same as that of the first embodiment, the description thereof is omitted. In addition, when the phrase in the text given by the user matches the headword in the dictionary database 31, the proficiency learning unit 23 sets the paraphrase unnecessary in the user information storage table 32, and the phrase becomes a paraphrase. If they match, a function to set that paraphrase is necessary is provided.
[0039]
<motion>
FIG. 10 is a flowchart illustrating the operation of the second embodiment.
Also in the specific example 2, the user performs a user authentication process by a method such as inputting a user ID from the input unit 11 (step S201), and selects whether to learn the proficiency level or perform a paraphrase process (step S201). S202). In step S202, in the case of "Y", the operation of the conversion unit 22 is performed, and in the case of "N", the operation of the proficiency learning unit 23 is performed.
[0040]
1) When learning the proficiency (in the case of "N" in step S202)
When the user inputs a text from the input unit 11, the proficiency learning unit 23 performs a proficiency learning process (Step S203). Here, the text input by the user may be a text created by the user or a web page on the Internet. In other words, terms used in texts created by the user and texts that the user can read and understand can be considered to be words acquired by the user. From such a viewpoint, the proficiency learning process of the target user is performed.
[0041]
FIG. 11 is a flowchart illustrating the flow of the proficiency learning process.
When performing the proficiency learning process, first, the input text is subjected to morphological analysis (step S221). Next, the words of the morphological analysis result are searched in a dictionary one by one, and it is checked whether or not the words are registered in the dictionary database 31 (steps S222 and S223). If the word is registered in the paraphrase storage unit 312 of the dictionary database 31, information “O” indicating the need for paraphrase is stored in a corresponding column of the user information storage table 32 (steps S224 and S225). Shift to the processing of the word (step S226). On the other hand, if the word is registered in the headword storage unit 311, information “x” indicating that paraphrase is unnecessary is stored in a corresponding column of the user information storage table 32 (steps S 224 and S 225), and the next word is stored. (Step S226). If it is not registered in any of the headword storage unit 311 and the paraphrase storage unit 312 in the dictionary database 31, the process proceeds to the next word without performing any processing (steps S224 and S226). When the processing of steps S211 to S225 is completed for all the words in the input text, the proficiency learning processing ends.
[0042]
In addition, for data in which the paraphrasing / unnecessary information in the user information storage table 32 is blank, according to a predefined rule (for example, “O” or “X” is collectively assigned according to the level of the condition storage unit 313). ) Paraphrasing / unnecessary information may be added.
[0043]
2) When performing paraphrase processing (in the case of "Y" in step S202)
In the specific example 2, the paraphrasing processing of steps S204 to S206 is the same as the processing of steps S105 to S107 in the specific example 1. Then, in the specific example 2, after the paraphrase processing result is displayed on the output unit 12, if the user wants to learn using the result (step S207), the processing result is corrected (step S208). For example, when the user's proficiency increases, a correction is made such that the paraphrased word is returned to the original word (headword). In such a case, it is desirable to write the original word and the paraphrase in the output result.
[0044]
Then, when the user wants to perform proficiency learning using the corrected text, the user inputs this text as a proficiency learning text from the input unit 11. Thereby, the proficiency learning unit 23 performs the process shown in FIG. 11 and reflects the process on the user information storage table 32.
[0045]
Thereafter, it is determined whether to process a new text (step S209). If the new text is to be processed, the process returns to step S204. If not, the user logs out (step S210) and ends the process.
[0046]
<effect>
As described above, according to the specific example 2, when a word in a document for learning proficiency matches a headword or paraphrase in the dictionary database 31, the proficiency learning unit 23 does not need to paraphrase the word. Since the information necessary for paraphrase is reflected in the user information storage table 32, the user does not need to register the necessity of word paraphrase one by one. Can be easily customized.
[0047]
Further, since the paraphrase processing result is corrected and the learning of the proficiency level is performed, the user information storage table 32 can be easily and reliably customized.
[0048]
<< Specific Example 3 >>
In Example 3, the paraphrasing system is configured by a client (input / output device) and a server (processing device, database).
[0049]
<Constitution>
FIG. 12 is a configuration diagram of the third embodiment.
The illustrated system is realized by connecting a client 100 and a server 200 via a communication line. The input / output device 10a is provided on the client 100 side, and the processing device 20b and the storage device 30a are provided on the server 200 side. That is, the overall configuration of this example is substantially the same as in Examples 1 and 2, except that text is exchanged between the client 100 and the server 200 via the network.
[0050]
12, the input / output device 10a includes an input unit 11, an output unit 12, and a transmission / reception unit 13. Here, the input unit 11 and the output unit 12 are the same as those in the first and second embodiments. The transmission / reception unit 13 is a transmission / reception unit on the client 100 side for exchanging data with the server 200.
[0051]
The processing device 20b on the server 200 side includes a conversion unit 22 and a transmission / reception unit 24, and the conversion unit 22 has the same configuration as the first and second embodiments. The transmission / reception unit 24 performs transmission / reception on the server 200 side, such as receiving the text transmitted from the client 100 and transmitting the text to the morphological analysis unit 221 or transmitting the paraphrase processing result in the text conversion unit 223 to the client 100. It is a functional part for performing.
[0052]
The storage device 30a has the same basic configuration as those of the first and second embodiments, but differs in that it has only the dictionary database 31.
[0053]
<motion>
FIG. 13 is a flowchart illustrating the operation of the third embodiment.
When performing the paraphrasing process, the user transmits the target text and information on his / her proficiency level to the server 200 (step S301). Here, the user's own proficiency is, for example, information such as level 1 and level 2.
[0054]
In the server 200, when the transmission / reception unit 24 receives such information, the conversion unit 22 performs a paraphrase process (step S302). This paraphrasing process will be described later. When the paraphrasing process is performed in step S302, the transmission / reception unit 24 transmits the processing result to the client 100 (step S303). Thereby, on the client 100 side, the processing result is displayed on the screen by the output unit 12 (step S304). If there is a new text (step S305), the process returns to step S301 and repeats the above-described processing. Otherwise, the paraphrasing processing ends.
[0055]
FIG. 14 is a flowchart illustrating the operation of the paraphrasing process.
The flow of the paraphrase process is almost the same as in the first and second embodiments, but the user's proficiency information is input and transmitted on the client 100 side. The difference is that the necessity / unnecessity of paraphrase is determined by comparing the information of the storage unit 313.
[0056]
First, when the user's proficiency level information is received together with the text, the morphological analysis unit 221 performs morphological analysis on the text (step S311). Next, the dictionary search unit 222 extracts a word at a time from the morphological analysis result and performs a dictionary search of the dictionary database 31 (step S313). If the word is registered in the dictionary, the user's proficiency level is compared with the difficulty level registered in the dictionary database 31 (information in the condition storage unit 313), and the difficulty of the word is determined by the user. It is determined whether the proficiency level is higher than the proficiency level (step S314). If this is the case in step S314, the paraphrase is stored in the buffer (step S315). On the other hand, if the word is not registered in the dictionary database 31 or the difficulty of the word is lower than the user's proficiency, the word is stored in the buffer as it is without performing paraphrase (step S316).
[0057]
Then, such processing for each word is repeated, and when the last word ends (step S317), the paraphrasing processing ends.
[0058]
<effect>
As described above, according to the third embodiment, the processing device 20b for performing the paraphrasing process and the dictionary database 31 are placed on the server 200, and the text and the proficiency information are provided from the client 100. And the storage capacity of the client 100 side can be reduced. As a result, the paraphrasing processing system can be incorporated into an apparatus having a limited processing performance and storage capacity, such as a portable terminal.
[0059]
Note that, in the specific example 3, as in the specific examples 1 and 2, the user information storage table 32 may be provided on the server 200 side, and the client 100 may transmit identification information such as a user ID.
[0060]
<< Specific Example 4 >>
In Example 4, a paraphrase corresponding to a condition based on the contextual relationship of a headword is stored in a dictionary database, and when performing paraphrase processing, a paraphrase corresponding to the condition is selected. .
[0061]
<Constitution>
FIG. 15 is a configuration diagram of the specific example 4.
The illustrated system includes an input / output device 10, a processing device 20c, and a storage device 30a. Here, the basic configurations of the input / output device 10 and the storage device 30 are the same as those of the first and second embodiments, and thus the description thereof is omitted. The conversion unit 25 in the processing device 20c includes a morphological analysis unit 221, a dictionary search unit 222, a text conversion unit 223, and a condition matching unit 224. Although the processing device 20c includes the paraphrase information setting unit 21 of the specific example 1 or the proficiency learning unit 23 of the specific example 2, the illustration is omitted.
[0062]
The conversion unit 25 has a function of converting a paraphrase according to the condition of a headword, and the morphological analysis unit 221 to the text conversion unit 223 are the same as those in the first and second specific examples. The condition matching unit 224 has a function of selecting a paraphrase based on first and second conditions described later in the dictionary database 31a and passing the selected information to the text conversion unit 223.
[0063]
The dictionary database 31a includes a dictionary database 31a and a user information storage table 32. The user information storage table 32 is similar to the first and second examples. The dictionary database 31a is a database provided with paraphrases corresponding to different conditions for headwords, and is configured as follows.
[0064]
FIG. 16 is an explanatory diagram of the dictionary database 31a of the specific example 4.
As shown, the dictionary database 31 of the specific example 4 includes a headword storage unit 311, a paraphrase word storage unit 312, an attribute information storage unit 314, a first condition storage unit 315, and a second condition storage unit 316. Here, the headword storage unit 311 and the paraphrase storage unit 312 have the same configuration as the dictionary database 31 of the first to third specific examples. The attribute information storage unit 314 is a function unit that stores information indicating the meaning of the headword or paraphrase or the relation in context. The first condition storage unit 315 is a functional unit that stores the first condition, which is the same information as the condition storage unit 313 in the first to third examples. Further, the second condition storage unit 316 stores a condition based on the contextual relationship of the headword, that is, a second condition that is information indicating under what condition the headword is used. It is. For example, D161 in the figure means that the meaning of "confit" is "cooking method", and the condition for rephrasing "pickled in pickles" is that the input is the expression "configuration of ~" Is a "vegetable".
[0065]
<motion>
FIG. 17 is a flowchart illustrating the operation of the paraphrase processing of the fourth example.
FIG. 18 is an explanatory diagram illustrating an example of the input text.
In the specific example 4, application conditions are checked when performing a dictionary search, and paraphrase processing is performed only when the conditions are matched (steps S401 to S407). Hereinafter, a case where the text shown in FIG. 18 is input will be described as an example.
[0066]
First, when the text of the TX 181 in FIG. 18 is input, the morphological analysis is performed by the morphological analysis unit 221 (step S401), and the result is dictionary-searched one by one by the dictionary search unit 222 (steps S402 and S403). Referring to "Poiblon" in the dictionary database 31a shown in FIG. 16, since the second condition storage section 316 is empty, the paraphrase "Pepper" is stored in the buffer (steps S404 and S405). Next, since "no" is not registered in the dictionary database 31a, the word "no" is stored in the buffer as it is (steps S404 and S406).
[0067]
Further, when "Confid" is searched (step S403), there are two types of paraphrase candidates, D161 and D162, and the second condition for applying this data is described in the second condition storage unit 316. . In the case of TX181, since the word preceding "Confi" is "Poireblon", and the meaning of the word is "vegetable" from the information in the attribute information storage unit 314 of D163, D161 is applied. "Pickled" is stored in the buffer (steps S404, S405).
[0068]
Similarly, when the text of TX182 is input, the word preceding “Confi” is “anana”, and the meaning of the word is “fruit” from the information in the attribute information storage unit 314 of D164. Therefore, D162 is applied, and the paraphrase "candied" is stored in the buffer (steps S404 and S405).
[0069]
FIG. 19 is an explanatory diagram showing the contents of the buffer after performing the paraphrasing process.
As shown, in TX191, "confi" is paraphrased as "pickled", while in TX192, "confi" is paraphrased as "candied".
[0070]
When the paraphrasing process is completed, the contents of the buffer are displayed on the output screen in the same manner as in the first and second examples, and if there is a new text paraphrasing request, the same process is repeated. If there is no request, the process ends.
[0071]
<effect>
As described above, according to the specific example 4, the paraphrase is selected under a different condition for a certain headword, so if there is a possibility of a plurality of paraphrases, the condition described in the dictionary database 31a is used. , An optimal paraphrase can be selected.
[0072]
<< Specific Example 5 >>
In Example 5, the paraphrasing process is performed without performing morphological analysis on the text.
[0073]
<Constitution>
FIG. 20 is a configuration diagram of the specific example 5.
The illustrated system includes an input / output device 10, a processing device 20d, and a storage device 30. Here, since the configurations of the input / output device 10 and the storage device 30 are the same as those of the first and second embodiments, the description thereof will be omitted. The processing device 20d includes a paraphrase information setting unit 21 and a conversion unit 26. The paraphrase information setting unit 21 has the same configuration as that of the first and second specific examples. The conversion unit 26 includes a dictionary search unit 222a and a text conversion unit 223, and the text conversion unit 223 is the same as the text conversion unit 223 of each specific example. That is, the conversion unit 26 of the specific example 5 is different from the conversion unit 22 of the specific example 1 in that the morphological analysis unit 221 is not provided. In addition, the dictionary search unit 222a checks each character from the beginning of the vocabulary in the input text with the headword in the dictionary database 31 and, if there is a matching character string, performs the paraphrase processing of the headword. It is configured.
[0074]
<motion>
The flow of the processing is almost the same as that of the first embodiment, but the details of the paraphrasing processing are different.
FIG. 21 is a flowchart illustrating the operation of the paraphrasing process of the specific example 5.
In the specific example 5, the dictionary search unit 222a of the conversion unit 26 performs a dictionary search for a character string of m characters from the first character of the input text (steps S501, S502, and S503). For example, using the sentence TX51 in FIG. 5 as an example, a dictionary search is performed using a character string starting from the first character as a key, such as “Chi”, “Cho”, “Chori”, etc. (step S503). If there is a matching character string (step S504), paraphrase processing is performed (steps S505, S506, S507). At this time, the number of searches may be reduced by setting the longest value of the search character string. When a plurality of character strings are matched, the processing is performed by performing a paraphrasing process on all the character strings or selecting the longest character string among the matched character strings.
[0075]
In the example of TX51, since "chorizo" matches, "dry sausage" which is a paraphrase of "chorizo" is stored in the buffer (step S506), and n = 1 + 4 (step S507). In step S512, since n is not the last character, the process returns to step S502, and a dictionary search is performed from the fifth character “ON” (step S503). Since there is no headword starting with "ON" in the dictionary (steps S504, S508, S509), "ON" is stored in the buffer (step S510). Then, n = n + 1 (step S511). If the character is not the last character (step S512), the process returns to step S502.
[0076]
That is, in step S504, if the headword does not match, the search target is increased by one character from the nth character, and the search is performed up to the last character. If the headword still does not match, the nth character is not searched. Is stored in the buffer.
[0077]
When "ON" is stored in the buffer, a dictionary search for a headword starting with "RI" is performed (steps S502, S503, S508, S509). Since this is also not in the dictionary, "RI" is stored in the buffer (step S510).
The above processing is repeated until the end of the input text.
[0078]
<effect>
As described above, according to the specific example 5, although the accuracy of word recognition is lower than that of the specific example 1, the morphological analysis is not performed, so that there is an effect that the processing is reduced.
[0079]
《Usage form》
In each of the above specific examples, a restaurant menu has been described as an example of the text to be paraphrased, but the present invention is not limited to this, and is applicable to various fields only by changing the contents of the dictionary database 31 (31a). Can be. For example, the present invention can be applied to a system for changing a kanji to hiragana or a reading kana, or a system for rephrasing a hospital chart so that a patient can easily understand it.
In addition, by registering foreign words (Katakana) and Japanese (Kanji, Hiragana) or standard languages and dialects in association, they can be incorporated into a part of the document creation system or document proofreading system to express expressions and terms. The system can be unified.
[0080]
In each of the above specific examples, the display form of the paraphrase is described as a method of replacing the headword with the headword. However, the headword and the paraphrase may be described together, or the paraphrase is displayed when the headword is designated with a mouse or the like. May be displayed.
In addition, a different display may be performed so that the paraphrase is displayed with an underline or the display color is changed so as to be distinguished from a non-paraphrase.
[0081]
In each specific example, an example in which words are paraphrased has been described.However, when performing a dictionary search, if a search with multiple words can be performed, even if the target of paraphrase is a word composed of a plurality of words, such as idioms and idioms I do not care.
[0082]
In each specific example, the dictionary database 31 (31a) and the user information storage table 32 are provided separately, but these may be provided as an integrated database. The user information storage table 32 may be held for each user.
[0083]
In each specific example, the condition stored in the condition storage unit 313 or the first condition storage unit 315 has been described as a one-dimensional value. However, this condition is a multidimensional one obtained by combining two or more conditions. Is also good. For example, one of the conditions describes the difficulty level, and the other describes the notation of the paraphrase (such as hiragana, katakana, or kanji). Thus, when paraphrasing the headword, the user 1 can prioritize the paraphrase in katakana, while the user 2 can perform processing such as prioritizing Japanese (kanji).
[0084]
In the specific example 2, the method of registering the paraphrase information for each word interactively using the paraphrase processing result has been described. However, a means for registering the necessary information collectively from a file in which necessary information is described is provided. You may.
When the user corrects the result of the paraphrasing process, a monitoring unit for this process may be provided to acquire information on a word that the user wants to paraphrase or a word that the user does not want to paraphrase.
[0085]
In the specific example 4, an example in which the semantic information of the word is stored in the attribute information storage unit 314 is shown. However, the present invention is not limited to this. Various information can be stored.
[0086]
In the specific example 5, the configuration of the dictionary database 31 of the specific examples 1 to 3 was used as the configuration of the dictionary database 31. Is also good.
[0087]
【The invention's effect】
As described above, according to the present invention, for a headword, a dictionary database for storing a paraphrase for expressing the headword in another expression is provided, and an arbitrary phrase is converted to a paraphrase using this dictionary database. As such, it is possible to rephrase difficult-to-understand expressions such as technical terms into easy-to-understand expressions.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a specific example 1 of a paraphrasing system of the present invention.
FIG. 2 is an explanatory diagram of a dictionary database and a user information storage table of a specific example 1.
FIG. 3 is a flowchart illustrating an operation of a specific example 1.
FIG. 4 is a flowchart illustrating an operation of a paraphrasing process of a specific example 1.
FIG. 5 is an explanatory diagram showing a restaurant menu as an example of a document.
FIG. 6 is an explanatory diagram of a result of morphological analysis of one sentence of the text in FIG. 5;
FIG. 7 is an explanatory diagram of the contents of a work buffer.
FIG. 8 is an explanatory diagram of a menu after the paraphrasing process.
FIG. 9 is a configuration diagram of a specific example 2.
FIG. 10 is a flowchart illustrating an operation of a specific example 2;
FIG. 11 is a flowchart showing proficiency learning processing.
FIG. 12 is a configuration diagram of a specific example 3.
FIG. 13 is a flowchart illustrating an operation of a specific example 3.
FIG. 14 is a flowchart showing an operation of a paraphrase process.
FIG. 15 is a configuration diagram of a specific example 4.
FIG. 16 is an explanatory diagram of a dictionary database of a specific example 4.
FIG. 17 is a flowchart illustrating the operation of the paraphrasing process of Example 4;
FIG. 18 is an explanatory diagram of an input text of a specific example 4.
FIG. 19 is an explanatory diagram of the contents of a work buffer according to a specific example 4.
FIG. 20 is a configuration diagram of a specific example 5;
FIG. 21 is a flowchart showing the operation of the paraphrase processing of the specific example 5.
[Explanation of symbols]
10, 10a I / O device
21 Paraphrase information setting section
22, 25, 26 conversion unit
23 Proficiency Learning Department
31, 31a Dictionary database
32 User information storage table

Claims (10)

A dictionary database that associates the headword with a paraphrase that expresses the headword in another expression,
A paraphrase system comprising: a conversion unit that refers to the dictionary database for an arbitrary phrase and, when there is a matching headword, converts the phrase into a paraphrase and outputs the paraphrase. The paraphrasing system according to claim 1,
A user information storage table for each user indicating whether or not to paraphrase the headword,
A paraphrase system, comprising: a conversion unit that converts an arbitrary phrase into a paraphrase based on user information corresponding to the user when the user is designated. In the paraphrase system according to claim 2,
A paraphrase system comprising: a paraphrase information setting unit that, when a user specifies whether or not to paraphrase a headword, reflects the specified content in the user information storage table. In the paraphrase system according to claim 2 or 3,
A paraphrase characterized in that when the word in the document given by the user matches a headword in the dictionary database, the user information storage table includes a proficiency learning unit that sets the headword to be not paraphrased. system. In the paraphrase system according to claim 2 or 3,
If the word in the document given by the user matches the paraphrase in the dictionary database, the paraphrase characterized in that the user information storage table is provided with a proficiency learning unit that sets the headword as paraphrase required. system. In the paraphrase system according to any one of claims 1 to 5,
An input / output device connected to the dictionary database and the conversion unit via a communication line is provided,
A paraphrase system, wherein the input / output device is configured to transmit an arbitrary phrase together with a paraphrase request to the conversion unit, and to receive a paraphrase result from the conversion unit. In the paraphrase system according to any one of claims 1 to 6,
A dictionary database with paraphrases corresponding to conditions based on the contextual relationship of the headword,
For any phrase, based on information indicating the contextual relationship of the phrase, refer to the dictionary database, and if there is a headword whose information satisfies the condition based on the contextual relationship, A paraphrase system comprising a conversion unit for converting a paraphrase corresponding to a headword. In the paraphrase system according to any one of claims 1 to 7,
A paraphrase system comprising: a conversion unit configured to morphologically analyze an input character string to extract words and convert a headword into a paraphrase based on the extracted words. In the paraphrase system according to any one of claims 1 to 7,
A character string is added one character at a time from an arbitrary character in the document, the character string is compared with a headword, and when a match is found, a conversion unit that determines that the character string is a phrase to be paraphrased is provided. Paraphrase system characterized by the following. In the paraphrase system according to any one of claims 1 to 9,
A paraphrase system characterized in that a portion converted into a paraphrase is displayed differently from a portion not converted.

Images