JPH01240924A - Negative expression corresponding japanese language analyzing system - Google Patents

Abstract

PURPOSE:To generate an enquiry text based on more natural idea by providing a means to recognize negative expression and a negative range. CONSTITUTION:When a Japanese language character string written by each clause is inputted, a keyword qualification part 2 is started up, and a keyword is qualified by referring to a keyword directionary 1 at every clause, and keyword information is stored in a clause storage area 3. A negative expression qualification part 5 collates the clause storage area 3 in which the keyword information is stored with a negative word table 4, and negative expression is recognized, and negative information is stored in the clause storage are 3. A qualification part 6 for the negative range refers to the clause storage are 3 for which the negative expression is qualified, and sets negative range information on the clause storage area 3, and outputs the content of the information. In such a way, more natural conversational environment can be offered to a user by utilizing the negative expression in an input Japanese language character string, and a load on the user at the time of generating the enquiry text can be reduced.

Description

[Detailed description of the invention]

[Industrial Application Field] The present invention relates to a Japanese language analysis section of a database search system or the like having a Japanese language processing function, and is suitable for recognizing negative expressions from input H frozen character strings and analyzing their range. Concerning Japanese language analysis methods. [Prior Art] As described in Japanese Patent Application Laid-Open No. 58-88971, the conventional Japanese language analysis method extracts independent words from an input Japanese character string and determines the semantic relationship between the independent words. It was an attempt to analyze. However, using information from independent words, it was not possible to recognize negative expressions, and it was not possible to use negative expressions in query sentences. [Problems to be Solved by the Invention] The above-mentioned prior art does not take into consideration the case where a negative expression is included in an inquiry sentence, and there is a problem in that a negative expression cannot be used in an inquiry sentence. The purpose of the present invention is to provide the user with a more natural conversation environment and reduce the burden on the user when creating a query sentence by making it possible to use negative expressions in the input H frozen character string. It's about doing. [Means for resolving the problem] The above purpose is to use the Japanese language analysis department to manually

[After processing at position a and storing each clause in a storage device in a format consisting of a clause character string and clause information, a means for recognizing a negative expression by referring to the clause character string, and This is achieved by providing means for recognizing the range of negation by referring to the phrase information after recognizing the expression. [Effect]

The means for recognizing negative expressions compares each phrase with a negative word table, and when a character string in the negative word table is included in the phrase character string, recognizes that the phrase is negated. The means for recognizing the negative range checks the clause information 2 of the clause ahead of the clause in which the negative expression has been recognized, and recognizes the range in which it satisfies the conditions as the negative range. [Example] An example of the present invention will be described below with reference to FIGS. 1 to 11. FIG. 1 shows the functional blocks of the F1 ice-packing analysis method using the present invention, in which 1 is a keyword dictionary, 2 is a keyword recognition section, 3 is a phrase storage area, 4 is a negative word table, and 5 is the recognition part of the negative expression, and 6 is the recognition part of the range of negation. When the written H iced character string is input from Bunsetsu,
The keyword recognition section 2 is activated, the keyword dictionary 1 is referred to for each phrase, the keyword is recognized, and the keyword information is stored in the phrase storage area 3. The negative expression recognition unit 5 compares the phrase storage area 3 in which keyword information is already stored with the negative word table 4, recognizes a negative expression, and stores the negative information in the phrase storage area 3. 6 refers to the assessment table@certified phrase storage area 3, sets negative range information in the phrase storage area 3, and outputs this content. The functional blocks shown in FIG. 1 will be described in detail from FIG. 2 onwards. FIG. 2 shows the hardware configuration of an embodiment of the present invention, in which 7 is a processor, 8 is a display device for outputting the negative range recognition result, and 9 is a keyboard input device for inputting H ice string. , 1o is a storage device, which consists of the following storage units. 101 is a program storage area, 11 is a keyword dictionary storage area, 41 is a negative word table storage area,
102 is a human-powered Japanese character string storage area, 3 is a bunsetsu storage area, and 103 is a work area. FIG. 3 is a flowchart showing the contents of the keyword recognition section. When the keyword recognition section starts, it will try to obtain the written H iced string from the bunsetsu (21), refer to the first bunsetsu (22), determine whether the bunsetsu string is stored (2.3), and if it is not stored. , the process ends, and if it is stored, branches to 24. At step 24, it is determined whether the keyword exists in the phrase by referring to the dictionary, and if the keyword exists, the process branches to 25, and if it does not exist, the process branches to 26. In step 25, the phrase is divided into keywords and attached parts and set in the phrase storage area. , 26, the process moves to the next clause and branches to 23. FIG. 4 shows an example of an input Japanese character string, which is the human input shown in FIG. FIG. 5 shows a keyword dictionary format, which is composed of a keyword, an item code, and a type code, and shows an example corresponding to the H ice-packed character string shown in FIG. 4. FIG. 6 shows a one-phrase storage area. This consists of a keyword, an appendix 2 item code, a type code, and negative information, and is the output obtained by processing the human-powered Japanese character string in FIG. 4 with the keyword recognition section in FIG. 3. FIG. 7 is a flowchart showing the contents of the negative expression recognition section. When the negative expression recognition unit is activated, it refers to the phrase storage area at the beginning (51) and determines whether a keyword is stored (52). If it is not stored, the process ends; if it is stored, it branches to step 53. . At step 53, the appendix of the bunsetsu storage area is referred to, and at step 54, by checking with the negative word table, it is determined whether a negative expression is included. If it is included, the process branches to 55, and if it is not included, the process branches to 56. At 55, a negation symbol is set in the negation information in the phrase storage area, and the process branches to 56. On the 5th day, the process advances to the next clause and branches to 52. FIG. 8 is a negative word table, which is composed of negative word headings, and shows an example corresponding to the Japanese character string shown in FIG. FIG. 9 shows the state of the clause storage area after the negative expression recognition process has been executed, and is the output obtained by inputting the clause storage area of FIG. 6. FIG. 10 is a flowchart showing the contents of the negative range recognition section. When the negative range recognition unit starts up, it refers to the first phrase in the phrase storage area and checks 60.1. ), it is determined whether the keyword is stored (602) - If the keyword is not stored, the process ends Y; if it is stored, the process branches to 603. In 603, the negation information in the bunsetsu storage area is referred to, and in 604 it is determined whether a negation symbol is set.
If not set; h, branch to 613; if set, branch to 605. At 605, the location of the phrase storage area currently being processed is memorized (the location is designated as P), and the process branches to 606. In 606, it is determined whether there is a keyword ahead, and if not, the process branches to 612.
Branch to 07. 607 refers to the preceding clause, and 6
At 08, it is determined whether the keyword type code is data or attribute. If it is an attribute, the process branches to 612, and if it is data, the process branches to 609. At 609, the item code of the keyword at the position of P is compared with the item code of the keyword currently being processed. At 610, it is determined whether the two item codes are equal. If they are not equal to 1, the process branches to 612. Branches to [61]. In 611, a negation symbol is set in the negation information area of the clause storage area corresponding to the keyword currently being processed, and in 6612, which branches to 606, the position of the clause storage area to be processed is set to P, and the process branches to 613. At 613, refer to the clause storage area at the next position, and at 60
Branch to 2. FIG. 11 shows the state of the phrase storage area after the processing of the negative range recognition unit has been executed, and is the output obtained as the phrase storage area gunman force shown in FIG. 9. The embodiments have been described in detail above. According to this embodiment, since each function can be configured into blocks, it is possible to easily incorporate the five negative expression recognition function and the main negative range recognition function into a conventional Japanese language analysis section. In addition, since each functional block is realized by program logic, speed can be increased by converting each functional block to T and SI.Also, since the processing information to be used is stored separately, it can be used depending on the field of application. It has the effect of being able to be used for general purposes by replacing the information content. [Effects of the Invention] According to the present invention, since users can use negative expressions in inquiry sentences, they can create inquiry sentences based on more natural ideas, which reduces the burden of creating inquiry sentences and improves operability. effective.

[Brief explanation of the drawing]

Fig. 1 is a functional block diagram of the To1 ice pack analysis method using the present invention, Fig. 2 is a block diagram showing the hardware configuration of an embodiment of the present invention, and Fig. 3 is a block diagram of the keyword recognition section of this embodiment. Processing flowchart, FIG. 4 is an example of a Rokukado 1 Japanese character string, FIG. 5 is an explanatory diagram showing the storage format of the keyword dictionary, FIG. 6 is an explanatory diagram showing the bunsetsu storage area, and FIG. 7 is an illustration of the present invention FIG. 8 is an explanatory diagram showing the negative word table, FIG. 9 is an explanatory diagram showing the phrase storage area after execution of the negative expression recognition process, and FIG.
The figure is a processing flowchart of the negative range recognition section according to the embodiment of the present invention, and FIG. 11 is an explanatory diagram showing the clause storage area after the negative range recognition process is executed. 1...Keyword dictionary, 2...Keyword certification department,
3... Clause storage area, 4... Negative word table, 5
... Negative expression certification section, 6... Negative range certification section. No. / m q-Tongue to the self-sniffing Tiful S'JEj 禮jishi translator and 2 No. Prisoner No. 7 Zu Imo 3 Sen No. 2 Mine//Ka No. 10 Masa (A) No. /Q Lu

Claims (1)

[Scope of Claims] 1. A Japanese language analysis method for negative expressions, characterized in that a Japanese language analysis unit included in a database search system or the like is provided with means for recognizing negative expressions and the range of negation. 2. As a means of recognizing negative expressions and the scope of negation,
Item 1, comprising the step of processing the input Japanese character string in units of clauses and setting each clause in an area on the storage device that is composed of a clause string storage section and a clause information storage section. Japanese language analysis method for negative expressions in the description. 3. Negative expression support according to item 2, characterized in that the means for recognizing negative expressions includes the step of storing a negative headword in a storage device and comparing this with the clause character string storage unit. Japanese analysis method. 4. The Japanese language analysis method corresponding to negative expressions according to item 2, further comprising the step of referring to the clause information storage unit as means for recognizing the range of negation.