JPS6190271A - Input word search method - Google Patents

Abstract

PURPOSE:To display words that include a given spelling or display plural dictionary contents by inputting the spelling to the extent it is known, when the spelling is unclear during the input of sentence. CONSTITUTION:When there is a word for which only a part of spelling of it is known during the input of sentence, only the known part of spelling is inputted, and then a search key 4a of keyboard 4 is depressed. The word thus inputted is stored in a buffer, and the number of the word inputted is counted. Next, the first letter of the word in the buffer is compared with the word designated by the dictionary pointer which is set at the head of the dictionary, and in case when they are the same, the next letter is compared likewise. And when the second letters are the same, same comparison is made to the third letters. This procedure is repeated as many times as the number of the letter counted of the inputted word. And if all of them are the same, the words shown by the dictionary pointer are displayed on the CRT3. Thus, by inputting the spelling of the word to the extent they are known, the inputted word or the plural dictionary contents can be displayed.

Description

[Detailed Description of the Invention] Industrial Field of Application The present invention is applicable to an interactive translation system.
This invention relates to an input word search and display method for efficiently proceeding with text processing.

BACKGROUND TECHNOLOGY In a machine translation system, there are times when the spelling of an ilt word becomes unclear during the translation process.In the prior art, the system inputs the spelling as much as it remembers. If the translation fails, the user has to change the spelling and enter the i again. Alternatively, they would look up the correct spelling in a book, look up the correct spelling, and do it manually. In such prior art, it is unclear how many times the correct spelling will be discovered, and there is also the problem that the interactive feature of this system cannot be taken advantage of.

Problems to be Solved by the Invention The present invention solves the above-mentioned problems, and by inputting the known spelling of Rh for a word whose spelling is unknown while inputting an input sentence, It is an object of the present invention to provide a search and display method for input words in which words containing m-ri are displayed.

Means for Solving Problems The present invention provides an input word method, characterized in that by inputting characters of a part of a word to be translated, one or more dictionary contents including the part are displayed. This is a search display method.

Embodiment fIS1 FIG. 1 is a block diagram showing the configuration of a translation device in which the present invention can be implemented. In relation to the central processing unit (CPU) 1 that performs processing such as translation processing, the CPU
main memory 2 where programs etc. for processing are stored, a display vc (CRT>3) for displaying the displays necessary for the translation process, a keyboard 4 for inputting necessary for the translation process, and A translation module 5 is provided in which necessary information is stored.The keyboard 4 also includes a word search key 4a and a return key 4b.
etc. are provided.

Here, as shown in Figure 2, the dictionary lookup form! ! Consider a translation monoaire 5 that consists of four parts: an elementary analysis section, a syntactic analysis section, a language conversion section, and a target language generation section.

Next, general machine translation processing will be explained.

8! Machine translation generally has levels as shown in Figure 3. When the source language is input from the top left, ■ Dictionary lookup,
6 Machine translation, which involves morphological analysis (■) and syntactic analysis and analysis (■), can be broadly divided into two types depending on the level of analysis. One is a pivot method in which concepts (referred to as intermediate M words) that do not depend on each M word in ■ are analyzed and a target M word is generated from there. The other is called the Trance 77 one-way method, which includes ■morphological analysis, ■ syntactic analysis,
The internal structure of the source language is obtained by performing either the semantic analysis (2) or the context analysis (2), and then the target language is generated after converting it to the same level structure of the target language. Here, the contents of each analysis are listed in Table 1.

(The following is a blank space) Table 1 In the present invention, it is assumed that the syntax analysis level (at least up to the level of parsing (■)) can be improved.

FIG. 4 shows an embodiment of the translation module 5.

Translation monoale 5 has 5 buffers A%810%D
, E, and a table 6 having a translation dictionary, grammar rules, and tree structure conversion rules. These buffs 7rA-
E stores data at each analysis level in the translation process, as will be described later.

Figure 5 is! 'I' The contents of each pan 7y in the S4 diagram are
read a book.

This example shows how to translate an English sentence into Japanese. First, the original text is read into the buff 7A of FIG. 4 as shown in ft55 (1). Necessary information about each car ah is extracted by the dictionary lookup morphological analysis section shown in Pt52 and put into the buffer 7B shown in FIG. The part of speech information that is part of it is
th+-, as shown in Figure 5 (2), is a multi-part speech word, which is uniquely determined by the parsing unit shown in Figure 2 below, and the dependencies are the same as in Figure 4. Buffer CI: Entered as shown in FIG. 5 (3). At this time, table 6 in Figure 4
Based on the grammatical rules, the analysis shown in Table 2 is performed.

(Margins below) Table 2: In this rule, for example, in the three-word conversion section in Figure 2, which expresses that "a sentence consists of a subject and a predicate,"
Similarly to syntax analysis, the tree is also analyzed using 'JJ, till' of the seven structure transformation, and the result is put into the buffer rD of Figure 4 as shown in Figure 5 (4). / )
In the mouth aJ generation section, appropriate particles and auxiliary verbs are added, the form is changed to Japanese ari, and the result is placed in the buffer r lself in Figure 4 as shown in Figure 5 (5). This is the output of 1 Nagi Mononeal 5.

rj! 6 and 7 show -'Aa example 70- of the present invention.
FIG. 8 shows an example of dictionary contents of this system, and FIG. 9 shows an example of display on the CRT3.

First, input a word whose spelling is only partially known and press the word search key 4a on the keyboard 4.The input word is stored in the buffer at step n1.Next, step n2
The number of characters in the input word is counted, and this number of characters is taken as, for example, 鵬. Also, in step n3, the dictionary pointer is set to the beginning of the dictionary.

At step n5, the routine moves to the routine shown in FIG. 7 in which the word stored in the buffer 7 and the word indicated by the dictionary pointer are compared.

In step n1 of the ffS7 diagram, it is determined whether the first character of the word indicated by the dictionary pointer is the same as the first character of the word in the buff 77. If they are the same, step n
Go to step 2 and compare the second character; if not the same, go to step n4 of FIG. 6, the dictionary pointer is incremented by +1, and step n5 of FIG. Process. tp
If the characters compared in step n2 of diagram J7 are the same,
The process moves to step n3, and a judgment is made as to the identity of the third character fjS. If the characters compared in step n2 are not the same, the process moves to step n4 of FIG. 6, the dictionary pointer is incremented by +1, and the dictionary pointer is Processing from step n5 in FIG. 6 is performed for the next word.

In this way, the same judgment is made for the number m of characters of the input word counted in step 1)2 of Fig. 6, and if they are the same, the process moves to step n6 of Fig. 6 and the word indicated by the dictionary pointer is displayed on the CRT 3. do. Next, the process moves to step n7 of FIG. 6, where the dictionary pointer is incremented by 1. At step n8, the process moves to the routine of FIG. If the 1'JI cut at each step is "No", the process ends. If "Yes", the process proceeds to the next step sequentially, and if "Yes" at step nm in FIG. 7, the process ends.
Returning to step n6 in the figure, the words in the dictionary are displayed on the CRT 3. If "No" is determined in step 11m of FIG. 8, the process ends.

Here, it is assumed that the contents of the dictionary are as shown in the eighth figure. 9th
If the user inputs "ba" from the keyboard 4 and presses return 'kr-4b as in the display area A1 of FIG. 9(1), the search results in the display area B1 of FIG. 9(2) are displayed. Also,
b″ in the same manner as in the display area A2 of FIG. 9(2).
If you input , search results like the display area B2 in FIG. 9 will be displayed. In this way, I was able to enter the part of a word I only knew partially and easily display the words that contained this part.

Effects of the Invention As described above, according to the present invention, when only a part of the spelling of a given word is known, it is possible to input the known spelling in @ and display the words containing this spelling. did it.

[Brief explanation of drawings]

Figure fjS1 is a block diagram showing the configuration of the 11 translation device that can implement the present invention, Figure ffS2 is a diagram showing the specific structure of the translation module 5, and Figure 3 is for explaining the level of mechanical translation processing. , fjS4 is a block diagram showing a large example of n translation monoyl, i5 is a diagram showing the contents stored in each buffer A to E, and FIG. - Chart, Fig. 7 is the 70-chart of the routine of steps n5゜n8 in Fig. 6.
- An example of a display on a CRT 3 showing the results of processing according to a chart. 5... Translation monoyl, 6... Table, A, ~
, E°°“/777

Claims (1)

[Claims] An input word search and display method characterized in that, by inputting characters of a part of a word to be translated, one or more dictionary contents including that part are displayed.