Computer aided text generation method and system

• US 5,200,893 A
• Filed: 02/26/1990
• Issued: 04/06/1993
• Est. Priority Date: 02/27/1989
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd., Hitachi Microcomputer Engineering Limited (Hitachi, Ltd.)
• Original Assignee: Hitachi, Ltd., Hitachi Microcomputer Engineering Limited (Hitachi, Ltd.)
• Inventors: Maeda, Kazuaki, Ozawa, Kuniaki, Kinukawa, Hiroshi
• Primary Examiner(s): Envall, Jr., Roy N.
• Assistant Examiner(s): Chung, Xuong
• Application Number:
  US07/484,888
• Time in Patent Office: 1,135 Days
• Field of Search: 364/419, 395/902, 395/927, 395/934, 395/144, 395/145
• US Class Current: 715/235
• CPC Class Codes: G06F 40/174 Form filling; Merging
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/56 Natural language generation

Abstract

A computer aided text generation system and method provides functions of aiding a logical outline structure of a text, aiding the text generation using a conventional expression, aiding the text generation without using conventional expression, aiding the refinement of the generated text and aiding the collection of illustrative sentences. The present system permits generation of a finished text of high quality.

Claims

1. A computer system which includes a central processing unit, a memory, an input device and an output device, said memory storing a file of commonly used expressions and a rule for use in determining a logical outline structure of a text from first data representing classification of the text and second data representing nuance of the text, a computer aided text generation method comprising:

(a) a first step of inputting the first data of the text by said input device;

(b) a second step of inputting the second data of the text by said input device;

(c) a third step of obtaining a logical outline structure of the text corresponding to said input first and second data by referring to said rule stored in said memory;

(d) a fourth step of graphically outputting said logical outline structure obtained by said third step to said output device;

(e) a fifth step of inputting third data designating at least a scene and a topic contained in said output logical outline structure from said input device in correspondence to the output logical outline structure; and

(f) a sixth step of selecting a commonly used expression corresponding to said second data representing nuance of the text from said file of commonly used expressions in response to said inputted third data designating at least a scene and a topic contained in said output logical outline structure.

2. A computer aided text generation method according to claim 1, further comprising a step of repeating said fourth to sixth steps for each of a plurality of paragraphs of the text.

3. A computer aided text generation method using a computer system which comprises a central processing unit, a memory, an input device and an output device, comprising:

(a) a first step of setting a type of a text;

(b) a second step of setting nuance of the text;

(c) a third step of obtaining a logical outline structure diagram indicative of the logical outline structure of the text following the step type and nuance of the text by referring to a rule stored in said memory;

(d) a fourth step of outputting the logical outline structure diagram to said output device obtained by said third step;

(e) a fifth step of selecting an element indicative of the contents of the text from the output logical outline structure diagram;

(f) a sixth step of selecting a commonly used expression from a file of commonly used expressions provided in said memory in response to said selected element;

(g) a seventh step of outputting said selected commonly used expression to said output device;

(h) an eighth step of accepting a request for correcting the output commonly used expression;

(i) a ninth step of correcting the selected commonly used expression in response to contents to be corrected corresponding to said accepted request;

(j) a tenth step of determining whether there is an error in the commonly used expression corrected by said ninth step of referring to said rule provided in the memory;

(k) an eleventh step of outputting, when there is an error as a result of the determination, to said output device a comment on the error;

(i) a twelfth step of accepting a request for further correcting the corrected commonly used expression in response to said comment from said output device; and

(m) a thirteenth step of further correcting the corrected commonly used expression in response to contents to be corrected corresponding to the correcting request in said twelfth step.

4. A computer aided text generation method comprising:

(a) a first step of setting first data representing classification of a text;

(b) a second step of setting second data representing nuance of the text;

(c) a third step of obtaining a logical outline structure of the text corresponding to said input first and second data by referring to said rule stored in said memory;

(d) a fourth step of graphically outputting said logical outline structure obtained by said third step to said output device;

(e) a fifth step of inputting third data designating at least a scene and a topic contained in said output logical outline structure form said input device in correspondence to the output logical outline structure; and

(f) a sixth step of selecting a commonly used expression corresponding to said second data representing nuance of the text from said file of commonly used expressions in response to said inputted third data designating at least a scene and a topic contained in said output logical outline structure.

5. A computer aided text generation method according to claim 4, wherein said third step includes:

a step of determining the logical outline structure of the text for each of a plurality of paragraphs of the text.

6. A computer aided text generation method comprising:

(a) a first step of setting first data on a text;

(b) a second step of setting second data on the text; and

(c) a third step of determining a logical outline structure of the text corresponding to the set first and second data by referring to a rule to determine the logical outline;

wherein said first data on the text includes a text classification, which is divided into hierarchies, each having a text classification table, said first step including;

a step of sequentially browsing the text classification tables for the hierarchies, starting from the uppermost table in the hierarchies; and

a step of obtaining a text classification stored in the lowermost table of the hierarchical text classification tables.

7. A computer aided text generation method according to claim 6, wherein said step of browsing the hierarchical text classification tables sequentially repeats, starting from the uppermost table of the hierarchical text classification tables;

a step of outputting to an output device the contents of the respective text classification table; and

a step of inputting one text classification in the output text classification table from an input device.

8. A computer aided text generation method comprising:

(a) a first step of setting first data on a text;

(b) a second step of setting second data on the text; and

(c) a third step of determining a logical outline structure of the text corresponding to the set first and second data by referring to a rule to determine the logical outline structure of text;

wherein said second step includes;

a step of setting, as the second data of the text, at least one of a style and a length of the text and degree of emphasis.

9. A computer aided text generation method according to claim 8, further including:

a step of outputting to an output device information stepwise indicative of the degrees of items indicated by the second data of the text; and

a step of selecting one piece of information from the output information.

10. A computer aided text generation method comprising:

(a) a first step of setting first data on a text;

(b) a second step of setting second data on the text; and

(c) a third step of determining a logical outline structure of the text corresponding to the set first and second data by referring to a rule to determine the logical outline;

wherein a two-dimensional table is provided, the first dimension of the table corresponding to the first data of the text, the second dimension of the table corresponding to the second data of the text, the classification of a logical outline structure being set at a position where the first and second dimensions intersect, said third step comprising;

a step of browsing the first dimension with the first data on the set text;

a step of browsing the second dimension with the second data on the set text; and

a step of obtaining the classification of a logical outline structure at a position in the two-dimensional table indicated by the results of browsing the first and second dimensions.

11. A computer aided text generation method comprising:

(a) a first step of setting first data on a text;

(b) a second step of setting second data on the text; and

(c) a third step of determining a logical outline structure of the text corresponding to the set first and second data by referring to a rule to determine the logical outline;

further including;

a step of providing a logical outline structure diagram file indicative diagrammatically of a logical outline structure of the text;

a fourth step of obtaining from said logical outline structure diagram file a logical outline structure diagram corresponding to the logical outline structure determined in the third step; and

a fifth step of indicating said logical outline structure diagram in a graphical manner.

12. A computer aided text generation method according to claim 11, wherein

said logical outline structure diagram is indicative of first information on the text and a relationship among the first information.

13. A computer aided text generation method according to claim 12, wherein said first information on the text includes second information on the text and third information on a text containing details of the second information.

14. A computer aided text generation method according to claim 13, wherein said first information on the text includes elements indicative of the contents of the text, wherein said second information on the text includes a scene developed in the text, and wherein said third information includes a topic related to the scene developed in the text.

15. A computer aided text generation method according to claim 13, further including:

a sixth step of selecting from said second information a desired one to be described in the text;

a seventh step of selecting from said third information a desired one to be described in the text in correspondence to the selected second information when occasion demands;

an eighth step of displaying the selected second and third data in a visually distinguishable manner different from that used for display of the second or third information which has not been selected in said seventh step.

16. A computer aided text generation method according to claim 15, further including:

a ninth step of setting, when a plurality of pieces of second information are selected, a logical outline structure of a text on the selected plurality of pieces of second information in the sequence of selecting said plurality of pieces of second information; and

a tenth step of setting, when a plurality of pieces of third information are selected corresponding to one piece of the selected plurality of pieces of second information, the logical outline structure of the text, on the selected plurality of pieces of third information in the logical outline structure of the text on the second information, in the sequence of selecting the plurality of pieces of third information.

17. A computer aided text generation method, using a computer system which comprises a memory storing a file of commonly used expressions, comprising:

(a) a first step of setting elements indicative of the contents of a text; and

(b) a second step of selecting a commonly used expression from said file of commonly used expressions in response to said elements set in said first step;

wherein when a plurality of commonly used expressions are used in correspondence to one of elements indicative of the contents of the text, information indicative of a sequence of using the plurality of commonly used expressions is stored in said file of commonly used expressions in correspondence to each of said plurality of the commonly used expressions, the method further including;

a third step of aligning, when said plurality of commonly used expressions are selected in the second step, the selected plurality of commonly used expressions by referring to said information indicative of sequence of using the selected plurality of commonly used expressions.

18. A computer aided text generation method according to claim 17, further including

a fourth step of outputting said selected commonly used expression to an output device.

19. A computer aided text generation method according to claim 17, further including:

a first step of selecting, when there is no commonly used expressions in said file of commonly used expressions corresponding to one of elements indicative of the contents of the text, a commonly used expression from said file of commonly used expressions in response to an element indicative of contents close to those indicated by the one element.

20. A computer aided text generation method, using a computer system which comprises a central processing unit, a memory, an input device and an output device, said memory including a file of commonly used expressions, information on correctable portions of the commonly used expressions of the file and comments on the correctable portions, comprising:

(a) a first step of setting first information on a text;

(b) a second step of selecting a commonly used expression from the file of commonly used expressions in response to said first information;

(c) a third step of outputting the selected commonly used expression to said output device;

(d) a fourth step of determining whether correction to the output commonly used expression is required or not;

(e) a fifth step of outputting to said output device a comment corresponding to a correctable portion of the output commonly sued expression when correction is required as a result of the determination; and

(f) a sixth step of correcting the commonly used expression output to said output device in response to the comment.

21. A computer aided text generation method according to claim 20, wherein said third step includes:

a step of outputting information on a correctable portion, corresponding to the selected commonly used expression, in a visually distinguishable manner different from that used for outputting information on other than the correctable portions by referring to the information on the correctable portion.

22. A computer aided text generation method according to claim 21, including:

a step of outputting, when there are a plurality of such correctable portions outputted in the visually distinguishable manner and including overlapping portions, information on the overlapping portions in a further visually distinguishable manner.

23. A computer aided text generation method according to claim 20, wherein the file of commonly used expressions includes a commonly used expression where the correctable portions include a blank, said third step including:

a step of outputting, when the commonly used expression having the blank is selected, the blank of the selected commonly used expression in a visually distinguishable manner different from that used to output other than the blank by referring to information on the correctable portions.

24. A computer aided text generation method using a computer system which comprises a memory which stores a first dictionary where an expression of a string of words is stored in correspondence to a keyword which includes a first word determined in the expression of the string of words, and a table where a second predetermined word is stored, comprising:

(a) a first step of deleting the same word as that stored in said table from a string of words constituting an expression to be retrieved;

(b) a second step of browsing said first dictionary using the remaining words as a keyword as a result of the deletion;

(c) a third step of obtaining, when the keyword is detected using the remaining words as a result of the browsing, from said first dictionary an expression corresponding to the keyword; and

(d) a fourth step of using the expression obtained for said first dictionary as the expression to be retrieved;

wherein a second dictionary is provided which stores expressions, different from those in said first dictionary, in correspondence to a keyword which comprises a second word determined in the expressions different from those in said first dictionary, said second dictionary storing information, indicative of expressions in the first dictionary similar in contents to the respective expressions in said second dictionary, in correspondence to the respective expressions in said second dictionary, further comprising;

a fifth step of browsing said second dictionary using the remaining words as the keyword when the keyword cannot be detected using the remaining words as a result of the browsing of said first dictionary; and

a sixth step of obtaining, when the keyword has been detected using the remaining words as a result of the browsing, an expression corresponding to that to be retrieved from said first dictionary by referring to information indicative of an expression in said first dictionary corresponding to that which in turn corresponds to the detected keyword.

25. A computer aided text generation method, using a computer system which comprises a central processing unit, a memory, an input device and an output device, wherein said memory stores as rules a method of describing items in a predetermined format portion of a text in correspondence to the items, comprising:

(a) a first step of outputting to said output device information on description guidance for each of the described items;

(b) a second step of inputting from said input device to said system information on the contents to be described for each of the described items;

(c) a third step of checking the input contents by referring to the rules;

(d) a fourth step of outputting to said output device information on the describing method corresponding to an error, if any, in the input contents as a result of the checking;

(e) a fifth step of inputting from said input device to said system information on corrected contents;

(f) a sixth step of correcting the input contents in response to the corrected contents; and

(g) a seventh step of converting the input contents to a predetermined format of the text.

26. A computer aided text generation method according to claim 25, wherein the fourth step includes:

a step of indicating the input contents involving an error in a visually distinguishable manner different from that used for indicating other than the contents involving an error.

27. A computer aided text generation method according to claim 25, wherein the input contents expected in the described item are stored in said memory in correspondence to the described item, and wherein said third step comprises:

a step of checking the input contents by referring to the expected input contents.

28. A computer aided text generation method according to claim 27, further including:

a step of checking the input contents by using as a don'"'"'t care at least one character of string of characters constituting the expected input contents.

29. A computer aided text generation method wherein a rule includes information of usage of a plurality of predetermined words and a comment on said usage of said predetermined words, comprising:

(a) a first step of parsing a text in order to correct an error included in the text;

(b) a second step of obtaining information of a predetermined part of speech and a plurality of words having said predetermined part of speech from a result of the parsing;

(c) a third step of checking said words having said predetermined part of speech with said predetermined words in said rule;

(d) a fourth step of displaying said predetermined words in said rule together with said comment on said usage of said predetermined words when coincidence is detected as a result of the checking; and

(e) a fifth step of correcting the text, in response to said comment;

wherein said fourth step comprises;

a step of displaying words, of the text, corresponding to the predetermined words in the rule, in a visually distinguishable manner different from that used for displaying other than those words of the text.

30. A computer aided text generation method information of usage of a plurality of predetermined words and a comment on said usage of said predetermined words, comprising:

(a) a first step of parsing a text in order to correct an error included in the text;

(b) a second step of obtaining information of a predetermined part of speech and a plurality of words having said predetermined part of speech from a result of the parsing;

(c) a third step of checking said words having said predetermined part of speech with said predetermined words in said rule;

(d) a fourth step of displaying said predetermined words in said rule together with said comment on said usage of said predetermined words when coincidence is detected as a result of the checking; and

(e) a fifth step of correcting the text, in response to said comment

further including;

a step of handling the predetermined words in the rule as ones the usage of which is mistakable.

31. A computer aided text generation method according to claim 30, wherein when there is one method of correcting of mistakable words, said one method of correcting is included in the comment, and wherein said fifth step includes:

a step of correcting the text to be refined, in accordance with said one correcting method.

32. A computer aided text generation method according to claim 30, wherein when there is a plurality of methods of correcting mistakable words, those methods of correcting are included in the comment, and wherein said fifth step includes:

a step of selecting one of said plurality of correcting methods to thereby correct the text to be refined.

33. A computer aided text generation method, using a computer system which comprises a central processing unit, a memory, an input device and an output device to generate a text, using illustrative sentences, comprising:

(a) a first step of inputting to said system a text from which illustrative sentences are to be collected;

(b) a second step of outputting to said output device information on guidance to a method of collecting an illustrative sentence;

(c) a third step of inputting to said system information on the collection of the illustrative sentence in correspondence to the output information; and

(d) a fourth step of collecting illustrative sentences from the text in accordance with said input information on collection of the illustrative sentence.

34. A computer aided text generation method according to claim 33, wherein said information on the guidance to a method of collecting an illustrative sentence includes information on a method of collecting a plurality of illustrative sentences, and wherein said third step comprises:

a step of inputting to said system information on at least one method of collecting an illustrative sentence from said information on the method of collecting a plurality of illustrative sentences.

35. A computer aided text generation method according to claim 34, wherein said third step includes:

a step of selecting at least one method of collecting illustrative sentences from a method of collecting illustrative sentences directed to collocations, a method of collecting illustrative sentences directed to function words and a method of collecting illustrative sentences directed to designated words.

36. A computer aided text generation method according to claim 35, wherein when the method of collecting illustrative sentences directed to collocations is selected, said fourth step comprises:

a step of extracting a simple sentence from the text from which illustrative sentences are to be collected;

a step of extracting elements of the simple sentence from said simple sentence;

a step of extracting a collocation from the relation among said elements of the simple sentence; and

a step of collecting as an illustrative sentence the simple sentence containing the extracted collocation.

37. A computer aided text generation method according to claim 35, wherein when said method of collecting illustrative sentences directed to the function words is selected, said fourth step comprises:

a step of outputting information on designation of a function word;

a step of designating the function word in response to the information on designation of the function word;

a step of designating the percentage of the designated function word contained in a sentence to be collected as an illustrative sentence; and

a step of collecting as an illustrative sentence a sentence containing the designated function word in a percentage larger than the designated one in each of the sentences of the text from which illustrative sentences are to be collected.

38. A computer aided text generation method according to claim 37, wherein said information on the designation of the function word includes that on a predetermined part of speech and at least one word or phrase contained in the part of speech, and wherein said step of designating the function word comprises:

a step of adding and performing at least one of deleting a displayed part of speech a word or a phrase.

39. A computer aided text generation method according to claim 35, wherein the method of collecting illustrative sentences directed to the function word is selected, said fourth step comprises:

a step of collecting as an illustrative sentence a sentence which includes at least one designated word in each of the sentences of the text to be collected.

40. A computer aided text generation method, using a computer system which comprises a central processing unit, a memory, an input device and an output device to generate a text using illustrative sentences, comprising:

(a) a first step of inputting to said system another text other than said text from which illustrative sentences are to be collected;

(b) a second step of outputting to said output device information on guidance to a method of collecting an illustrative sentence;

(c) a third step of inputting to said system, information on collection of the illustrative sentence in correspondence to the output information;

(d) a fourth step of collecting the illustrative sentence from the another text in accordance with the information on collection of the illustrative sentence;

(e) a fifth step of outputting to said output device, information on the collected illustrative sentence obtained in said fourth step;

(f) a sixth step of accepting the designation of a file, wherein the illustrative sentence is to be stored, in correspondence to the collected illustrative sentence outputted to said output device; and

(g) a seventh step of storing the collected illustrative sentence in the designated file.

41. A computer aided text generation system comprising:

(a) a memory storing a rule for use in determining a logical outline structure of a text from first data representing classification of said text and second data representing nuance of the text and a file of commonly used expressions;

(b) an input device for inputting to said system first data and second data of the text;

(c) an output device; and

(d) a central processing unit connected to said memory, said input device and said output device, said central processing unit obtains a logical outline structure diagram indicative of the logical output structure of the text corresponding to said input first and second data of the text by referring to said rule, graphically outputs information of said logical outline structure diagram to said output device, inputs, from said input device to said system, third data designating at least a scene and a topic contained in said output logical outline structure in correspondence to said output information of the logical outline structure diagram, and selects a commonly used expression corresponding to said second data representing nuance of said text from said file of commonly used expressions in response to said inputted third data designating at least a scene and a topic contained in said output logical outline structure.

42. A computer aided text generation system comprising:

(a) a memory storing a rule to determine a logical outline structure of a text from the classification and nuance of the text, a file of commonly used expressions and a text defining rule;

(b) an input device for inputting to said system classification and nuance of the text;

(c) an output device; and

(d) a central processing unit connected to said memory, said input device and said output device, said central processing unit obtains a logical outline structure diagram of the text following the classification and nuance of the text by referring to a rule provided in said memory, outputs information of the logical outline structure diagram to said output device, selects an element indicative of the contents of the text from the output information of the logical outline structure diagram, selects a commonly used expression from said file of illustrative sentences in response to said input element, outputs the selected commonly used expression to said output device, inputs from said input device to said system a request for correcting the output commonly used expression, corrects the selected commonly used expression in response to corrected contents corresponding to a request for correcting, determines whether the corrected commonly used expression involves an error by referring to said text refining rule stored in said memory, outputs, if a result of the determination indicates the presence of an error, to the output device a comment corresponding to the error, inputs from said input device to said system a request for further correcting the corrected commonly used expression in response to said comment, and further corrects the corrected commonly used expression in response to corrected contents corresponding to said request for correcting.

43. A computer aided text generation system comprising:

a memory storing a rule for use in determining a logical outline structure of a text from first data representing classification of said text and second data representing nuance of said text and a logical outline structure diagram file to display the logical outline structure diagram;

(b) first means for setting said first data of said text;

(c) second means for setting said second data of said text;

(d) third means responsive to an output from said first and second means for determining a logical outline structure of the text by referring to said rule stored in said memory;

(e) fourth means responsive to the output of said third means for obtaining a logical outline structure diagram from said logical outline structure diagram file stored in said memory;

(f) fifth means for graphically displaying the logical outline structure diagram;

(g) sixth means for inputting third data designating at least a scene and a topic contained in said output logical outline structure from said fourth means in correspondence to the output logical outline structure; and

(h) seventh means for selecting a commonly used expression corresponding to said second data representing nuance of the text from a file of commonly used expressions in response to said inputted third data designating at least a scene and a topic contained in said output logical outline structure.

44. A computer aided text generation system comprising:

(a) a memory storing a file of commonly used expressions, information on correctable portions of the respective commonly used expression of the file and comments on the correctable portions;

(b) first means for setting information on a text;

(c) second means responsive to the output of said first means for selecting a commonly used expression from the file of commonly used expressions;

(d) third means for determining whether correction to the commonly used expression selected by said second means is required or not;

(e) an output device for outputting the commonly used expression selected by said second means and a comment corresponding to a correctable portion of the output commonly used expression when a result of the determination by said third means requires correction; and

(f) an input device responsive to said comment from said output means for inputting to said system, corrected contents of the commonly used expression selected by said second means.

45. A computer aided text generation system comprising:

(a) a memory storing as a rule, a method of describing items in a predetermined format portion in a text in correspondence to the items;

(b) an output device for outputting information on description guidance for each of the described items;

(c) an input device for inputting to said system, contents to be described corrected for each of the described items;

(d) first means for checking the described contents by referring to said rule;

(e) second means for outputting, when the output of said first means indicates an error involved in the described contents, a relevant one among said rule corresponding to the error to said output device;

(f) third means responsive to said corrected contents for correcting said described contents; and

(g) fourth means for converting said input contents to a predetermined format of the text.

46. A computer aided text generation system comprising:

(a) an output device for outputting information a guidance to a method or collecting illustrative sentences to be used for text generation;

(b) an input device for inputting to said system, a text from which illustrative sentences are to be collected and information necessary for collecting illustrative sentences in correspondence to output information from said output device; and

(c) means responsive to input information necessary for collecting illustrative sentences from said input device, for collecting illustrative sentences from the text.

47. A computer aided text generation system comprising:

(a) a memory storing a file which in turn stores illustrative sentences;

(b) an output device for outputting an illustrative sentence and information on guidance to a method of collecting illustrative sentences to be used for text generation;

(c) an input device for inputting to said system (1) a text from which illustrative sentences are to be collected, (2) information necessary for collecting illustrative sentences in correspondence to output information from said output device and (3) information on designation of the file in said memory for storing illustrative sentences;

(d) means responsive to input information necessary for collecting illustrative sentences from said input device, for collecting from the text from which illustrative sentences are to be collected illustrative sentences; and

(e) means for storing the illustrative sentences collected by said collecting means in a designated location of said file.

48. A word processor apparatus comprising:

a central processing unit connected to a memory, an input device and an output device;

said central processing unit executes at least two of a logical outline structure program for determining a logical outline structure of a text, a commonly used expression selecting program for selecting a commonly used expression corresponding to the logical outline structure, a commonly used expression correcting program for correcting the selected commonly used expression, a syntax guiding program for indicting a syntax necessary for generating a text, a usage guiding program for indicating usage necessary for generating a text, a text refinement program for refining the generated text, and an illustrative sentence collecting program for collecting illustrative sentences used for generating a text, said programs being stored in said memory, wherein when said central processing unit is executing one of said programs, said central processing unit is capable of executing another program during execution of said one program.

49. In a system including a central processing unit, a memory, an input device and an output device, wherein said memory stores a rule for use in determining a logical outline structure of a text of a first language from first data representing classification of said text and second data representing nuance of said text, and a file of commonly used expressions in said first language,

a machine translation method of translating a second language to the first language comprising:

(a) a first step of inputting from said input device to said central processing unit, first data of the text to the second language;

(b) a second step of inputting from said input device to said central processing unit second data of the text in the second language;

(c) a third step of obtaining a logical outline structure of the text in the second language corresponding to said input first and second data referring to said rule stored in said memory;

(d) a fourth step of graphically outputting information on said logical outline structure to said output device;

(e) a fifth step of inputting from said input device to said central processing unit, third data designating at least a scene and a topic contained in said output logical outline structure in the second language in correspondence to the output information on the logical outline structure; and

(f) a sixth step of selecting a commonly used expression of the first language corresponding to said second data representing nuance of said text from said file of commonly used expressions in response to the third data designating at least a scene and a topic inputted in said fifth step.

50. A machine translation method of translating a second language to a first language, using a computer system which comprises a central processing unit, a memory, an input device and an output device, wherein said memory stores a file of commonly used expressions in the first language, information on correctable portions of commonly used expressions in the file, and a comment in the second language on the correctable portions, comprising:

(a) a first step of setting first information on a text in the second language;

(b) a second step of selecting a commonly used expression in the first language from said file of commonly used expressions in response to the first information set in said first step;

(c) a third step of outputting the conventional expression selected in said second step to said output device;

(d) a fourth step of determining whether correction to the output commonly used expression is required or not;

(e) a fifth step of outputting to said output device a comment corresponding to a correctable portion of the output commonly used expression when correction is required as a result of the determination by said fourth step; and

(f) a sixth step of correcting, in response to said comment, the commonly used expression outputted to said output device in the first or second language.

51. A machine translation method of translating a second language to a first language, using a computer system which comprises a central processing unit, a memory, an input device and an output device, wherein said memory stores as a rule, a method of describing an item in a predetermined format portion in a text in the first language in correspondence to that item, comprising:

(a) a first step of outputting to said output device, information on description guidance for each of the described items in the first or second language;

(b) a second step of inputting from said input device to said system the contents described for each of the described items in the first or second language;

(c) a third step of checking the input contents by referring to said rule;

(d) a fourth step of outputting, when an error is detected in the input contents as a result of the checking in said third step, to said output device the method of describing corresponding to the error in the input contents in the first or second language;

(e) a fifth step of inputting corrected contents for the error in the first or second language from said input device to said system;

(f) a sixth step of correcting the input contents in response to the corrected contents; and

(g) a seventh step of converting the input contents to a predetermined format in the text of the first language.

52. A method of collecting illustrative sentences, using a computer system which comprises a central processing unit, a memory, an input device and an output device, comprising:

(a) a first step of inputting information on a text from which illustrative sentences are to be collected;

(b) a second step of outputting to said output device, information on guidance to the method of collecting illustrative sentences;

(c) a third step of inputting information on collection of illustrative sentences in correspondence to the output information; and

(d) a fourth step of collecting an illustrative sentence from the text, from which illustrative sentences are to be collected, in accordance with the input information on collection of the illustrative sentences from said first step.

53. A method of collecting illustrative sentences, using a computer system which comprises a central processing unit, a memory, an input device and an output device, comprising:

(a) a first step of inputting to said input device, information on a text from which illustrative sentences are to be collected;

(b) a second step of outputting to said output device, information on guidance to the method of collecting illustrative sentences;

(c) a third step of inputting to said input device, information on collection of illustrative sentences in correspondence to the output information;

(d) a fourth step of collecting an illustrative sentence from said text, from which illustrative sentences are to be collected, in accordance with the input information on collection of illustrative sentences from said first step;

(e) a fifth step of outputting information on the collected illustrative sentence to said output device;

(f) a sixth step of accepting designation of a file which stores information on the output illustrative sentence in correspondence to that illustrative sentence; and

(g) a seventh step of storing information on the collected illustrative sentences in said designated file.