JPS6231475A - Natural language processor - Google Patents

Abstract

PURPOSE:To improve the input efficiency of sentences by changing the grammar rules and dictionaries of natural languages for analysis with input of plural natural languages mixed and therefore attaining the quick and easy input of sentences. CONSTITUTION:A sentence supplied through a microphone 1 is recognized by a voice recognizing part 2 and held by a sentence holding part 3 to be judged. If a switch 4 is pushed in said judgement process, the procedure proceeds to a switching process for sentence analysis information. While the procedure proceeds to a sentence analysis processing if the switch 4 is not pushed. In the switching process for sentence analysis information, a sentence analysis information selecting part 7 is switched in a prescribed cycle and the next sentence analysis information held by a sentence analysis information holding part 8 is supplied to a sentence analysis part 5. When these processes are through, the procedure proceeds to sentence analysis processing to analyze the sentence held by the part 3 by a sentence analyzing part 5 via the part 7 and by means of the supplied sentence analysis information. The result of this analysis is delivered to a sentence analysis result holding part 6.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a natural language processing device, and particularly to a natural language processing device that processes information related to natural languages such as Japanese and English.

[Prior Art] In general, natural language processing devices usually assume that an input sentence is a specific natural language, and process the input sentence using the grammatical rules and dictionaries of that natural language. However, in this case, since one specific natural language is assumed as input, there is a drawback that processing cannot be performed when a plurality of natural languages are mixed and input.

[Problems to be Solved by the Invention] Therefore, an object of the present invention is to appropriately select grammar rules and dictionaries that match the input natural language when a mixture of multiple types of natural languages is input. An object of the present invention is to provide a natural language processing device that can process natural language using the following methods.

[Means for solving the problem] In order to achieve such an objective, the present invention includes a sentence holding means for holding one sentence, a sentence analysis means for analyzing the sentence held in the sentence holding means, The present invention is characterized by comprising an analysis information holding means for holding linguistic information used for sentence analysis, an analysis information selection means for selecting the analysis information, and a switching means for switching the analysis information.

[Operation] According to the present invention, when a plurality of natural languages are mixed and input, the natural language division rules and dictionaries are switched and analyzed, so it is possible to input sentences quickly and easily. I can do it,
Wheat input work can be done efficiently.

[Example] The invention will now be described in detail with reference to one drawing.

FIG. 1 shows an embodiment of the present invention, where l is a microphone for inputting a sentence by voice, 2 is a voice recognition unit for recognizing the voice input from the microphone 1, and 3 is a voice recognition unit for recognizing the voice input from the microphone 1.
1 is a sentence holding section for holding sentences recognized by the speech recognition section 2, and 4 is a switch 4 attached to the microphone 1 to control the sentence analysis information selection 11&7.

5 is a sentence analysis unit that analyzes the sentences held in the sentence holding unit 3, and includes a CPU that performs arithmetic processing and an ROI in which a control program according to the flowchart in FIG. 2 is stored; 6 is a sentence analysis unit A sentence analysis result holding section 5 holds the analysis results, and a sentence analysis information selection section 7 switches the sentence analysis information used for sentence analysis. Reference numeral 8 denotes a sentence analysis information holding unit that holds linguistic information such as grammatical rules and dictionaries necessary for natural language analysis for each language, such as 8J for Japanese and American languages and 8E for English. Reference numeral 9 denotes a display device that displays the data of the sentence analysis result holding section 6 and the data of the sentence holding section 3.

Next, the operation of the apparatus shown in FIG. 1 will be explained with reference to the operation flow shown in FIG. 2.

In FIG. 2, first, the process moves from "Start" to the judgment in step Sl. In step Sl, the microphone l
The input sentence is recognized by the speech recognition section 2, and it is determined whether or not it is held in the sentence holding section 3. If the sentence is held, the process moves to step S7 to judge. In step St, if one sentence is not held in the sentence holding unit 3, step 1 is repeated.

Next, in the judgment in step S2, if the switch 4 has been pressed, the process proceeds to step S3, which is a sentence analysis information switching process, and if the switch 4 has not been pressed, the process proceeds to step S4, which is a sentence analysis process.

In step S3, the one-sentence analysis information selection unit 7 is cyclically switched at a predetermined period to supply the next sentence analysis information held in the sentence analysis information holding unit 8 to the sentence analysis unit 5.

When such sentence analysis information switching processing in step S3 is completed, the process moves to sentence analysis processing in step S4.

In step S4, the sentence held in the sentence holding unit 3 is
Zero or more sentence analysis processes that perform sentence analysis processing in the sentence analysis section 5 using the sentence analysis information supplied via the sentence analysis information selection section 7, and output the results to the sentence analysis result holding section B. When the process is completed, the process returns to the judgment in step Sl.

In the above embodiment, the sentence analysis information is switched cyclically, but it is also possible to provide a switch corresponding to the sentence analysis information, such as a Japanese switch or an English switch, and directly specify it. The sentence analysis information may also be switched.

In addition, in the above embodiment, the case where the one-sentence analysis information is switched cyclically has been explained, but when only two types of sentence analysis information are required, the second sentence is switched only when the Sweep 2000 button is pressed. The button may be switched to the analysis information and connected to the first sentence analysis information when the button is not pressed.

Alternatively, in the above embodiment, the information for one sentence analysis has been described as being Japanese and English, but it may also be one Japanese spoken and written, or one standard language and dialect.

Further, in the above embodiment, the case where there are two pieces of sentence analysis information has been described, but it goes without saying that there may be three or more pieces of information.

Furthermore, in the above embodiment, a case has been described in which the switch 4 for selecting sentence analysis information is attached to the microphone 1, but in addition to this, for example, a dedicated keyboard may be separately provided.

Additionally, in the above embodiments, only the case where sentences are input through a microphone has been described, but it is of course possible to input sentences using a keyboard, a character recognition device, etc.

[Effects of the Invention] As described above, according to the present invention, when a plurality of natural languages are mixed and input, analysis is performed by switching the natural language division rules and dictionaries. You can enter sentences into the computer, and you can perform literary works well.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a natural language processing device showing an embodiment of the present invention, and FIG. 2 is a flowchart showing an example of a control procedure of the embodiment of FIG. ■...Microphone. 2... Speech recognition section, 3... Sentence holding section, 4... Switch, 5... Sentence analysis section, 6... Sentence analysis result holding section, 7... Sentence analysis information selection section , 8... Sentence analysis information holding unit.

Claims (1)

[Scope of Claims] Sentence holding means for holding sentences; sentence analysis means for analyzing sentences held in the sentence holding means; and analysis information holding means for holding linguistic information used in the sentence analysis. A natural language processing device comprising: an analysis information selection means for selecting the analysis information; and a switching means for switching the analysis information.