JPH07192000A - Machine translation method equipped with summary function - Google Patents

Abstract

PURPOSE:To easily reflect the layout information of the original sentence to the translated one by comparing the length of the sentence before and after the translation and summarizing the translated sentence as short as the original one when it becomes longer than the original one. CONSTITUTION:A control part 108 accepts the translation request from an input part 101 such as a keyboard and sends the original sentence to a translation part 105. The part 105 decides the grammatical role of the phrases comprising the input original sentence by performing the morpheme analysis and the syntax analysis and finds out the proper word from the translation dictionary based on the result. Then,and executing a drying operation, based on a result calculated by a time counting means and a variation rate calculating means. CONSTITUTION:A motor 18 and a heater 13 are connected to a microcomputer 25 through a motor driving circuit 26 and a heater driving circuit 27 integrated into a controller 21, the cloth quantity of an object to be dried is calculated from the time required until an output of a drying degree detecting device 20 becomes a certain prescribed value after the operation is started, and the cloth qger than the original.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a machine translation device for efficiently translating a first language into a second language.

[0002]

2. Description of the Related Art In recent years, attention has been focused on a machine translation device that automatically translates a sentence in a first language into a sentence in a second language using a computer system. In the conventional machine translation device, the grammatical role of the words constituting the input original sentence is determined by morphological analysis or syntactic analysis, the translated word is found from the translation dictionary based on the result, and this is combined according to a predetermined rule. The translated text is created by doing so.

Further, there has been developed a technique for a translated text translated by such a translation system to obtain a translation result while the layout of the original text is reflected and the image of the original text is retained (Japanese Patent Laid-Open No. 4-259057 and the like). However, the length of the sentence is often different before and after translation, and the layout of the original sentence may not be reflected as it is. In such a case, there is a problem that, for example, the original document has one page of A4 format material and the translated document has two pages. To solve this problem, a technique has been developed in which the size of characters, line spacing, and character spacing are changed to restore the original layout (Japanese Patent Laid-Open No. 60-20283). But,
In this method, sentences are often hard to read.

There is also a technique for summarizing sentences (Japanese Patent Application No. 05-186570, etc.). Although it is possible to summarize and shorten the translation result using this technology, in the past, since the translation result was once converted into a text base and then input into the summarization system, not only was the operation complicated, but the layout However, there were problems such as the difficulty of restoring and the difficulty of setting an appropriate summarization ratio. Therefore, it is not possible to obtain a satisfactory result simply by using the conventional machine translation device and the sentence summarizing device in combination.

On the other hand, a function for retrieving a desired document from a document database has already been developed. However, when the user obtains the necessary information from the document database, if the search result is not in his / her native language (including a language that the user can use as easily as his / her native language, the same applies below), read it. Takes time. For this reason, a method of presenting it to the user after translation after translation is also devised. However, in a database search, not only information that is really needed but also information that is not needed are searched, so there is a problem in that it takes a lot of effort to read all of them and determine whether they are necessary. There was a lot of waste because we had to translate all the unnecessary data.

[0006]

As described above, in the conventional machine translation method in the machine translation apparatus, the sentence length is often different before and after the translation, and the layout of the original sentence cannot be reflected as it is. was there. The present invention has been made to solve the above problems,
It is a first object of the present invention to provide a machine translation method having a function of summarizing a sentence resulting from a translation so that the sentence has the same length when the sentence becomes longer than the original sentence.

Further, in the machine translation method used in the conventional machine translation apparatus having the search function, there is a problem in that all unnecessary data must be translated, which is wasteful. In addition, the user has to read the entire document in order to determine whether or not the search result is necessary, which is a problem of poor efficiency. The present invention has been made to solve the above problems, and is capable of presenting a translation result of only a summarized portion or a structurally important portion of document data obtained by a search to a user. A second object is to provide a method.

[0008]

In order to achieve the first object, in the machine translation method according to the present invention (claim 1), a translation step for translating a sentence in a first language into a sentence in a second language. And a length comparison step for comparing the length of the sentence in the first language before translation with the length of the sentence in the second language that is the translation result, and the sentence in the second language that is the translation result of the first language is the original sentence. And a summarizing step of summarizing the sentence of the second language so as to have a length substantially the same as that of the sentence of the first language when the sentence becomes longer than the sentence of the first language.

On the other hand, in order to achieve the second object,
In the machine translation method according to the present invention (Claim 2), a search step for searching a document from a document database described in the first language, a summarization step for summarizing the sentences searched by this search step, and this summarization step. A translation step of translating the document summarized in the step into a second language.

Instead of the above configuration, a search step for searching a document from a document database described in a first language, a translation step for translating the text searched in this search step into a second language, And a summarizing step of summarizing the document translated in the translating step.

As another means for achieving the second object, a search step for searching a document from a document database written in the first language and a logic of the text searched in this search step are provided. It may be characterized by having a logical structure analysis step of analyzing a structure and a step of translating only a structurally important portion into a second language according to the logical structure analysis result of this logical structure analysis step.

[0012]

According to the machine translation method of the present invention (Claim 1),
The length of the sentence in the first language before translation is compared with the length of the sentence in the second language, which is the translation result, and when it becomes longer than the sentence in the first language, the sentence in the second language is changed to the first sentence. Since it is summarized so that it is almost the same length as a sentence in one language, the translation result can be made almost the same length as the original sentence. Therefore, the layout information of the original sentence can be easily reflected in the translated sentence.

According to the machine translation method of the present invention (claim 2), the sentences described in the first language retrieved from the document database are summarized and then translated into the second language.
In database searches other than the native language, the summary of search results can be read in the native language. Therefore, the document search can be performed efficiently and easily.

[0014]

Embodiments of the present invention will be described below with reference to the drawings. First, a first embodiment of the present invention will be described. FIG. 1 is a block diagram showing a schematic configuration of a machine translation device according to this embodiment. As shown in FIG. 1, the machine translation device includes an input unit 101 and a display unit 10.
2, a recording unit 103, an editing unit 104, a translation unit 105, a length comparison unit 106, a summarization unit 107, and a control unit 108.

The input unit 101 is for inputting data such as sentences and commands such as a translation request, and is usually a keyboard or the like. The display unit 102 is the input unit 1
This is for displaying data input from 01, translation results, etc., and a bitmap display or the like is used. The recording unit 103 is for recording the data input from the input unit 101, the translation result, and the like.

The editing unit 104 performs an editing process on the data input from the input unit 101 and the data recorded in the recording unit 103. The translation unit 105 determines the grammatical role of the words constituting the input original sentence by morphological analysis, syntactic analysis, etc., finds a translated word from the translation dictionary based on the result, and combines the translated words according to a predetermined rule. Create a translation.

The length comparison unit 106 compares the lengths of the original sentence and the translated sentence. The summarizing unit 107 summarizes the translated sentence into the length of the original sentence according to the length comparison result of the original sentence and the translated sentence by the length comparing unit 106 when the translated sentence is longer than the original sentence. Control unit 108
Is for controlling these units and transferring data.

FIG. 2 is a flowchart showing an operation example of the control unit 108 of the machine translation device according to this embodiment. First, when a translation request is issued, the control unit 108 transfers the original sentence A to the translation unit 105 and obtains a translation result B (step S20).
1). Next, the length comparison unit 106 compares the original sentence A and the translation result B.
To obtain the length ratio c (step S202).

Whether the length ratio c is greater than 1 (that is, whether the translated sentence is longer than the original sentence) is determined (step S203). If the length ratio c is greater than 1,
The length ratio c and the translation result B are transferred to the summarizing unit 107, and the translation result B is summarized into a length of 1 / c (step S20).
4). With the above operation, the translation result B ′ having the same length as the original sentence A can be obtained.

Here, in the length comparison in step S202, the entire sentence can be compared, or the comparison can be performed in logical structure units such as chapters and sections, and physical structure units such as paragraphs and pages. . In addition, when the maximum length of characters that can be entered is specified for tables and columns,
It can also be compared to its maximum length. As the unit of length, in addition to the number of characters of the text data, the number of bytes, the product of the average width of the characters of the language and the number of characters, or the like may be used. For example, when Japanese is translated into English, English characters (alphabet)
Is often expressed in half the width of Japanese characters (kana, kanji), so an English sentence can be said to be longer when the number of characters doubles or more.

Here, the sentence length adjusting process may include a method of changing what is written in full spelling into abbreviations. This method will be described by taking the case of translating the word "machine translation" into English. In FIG. 3, the word "machine translation" appears in the original text in Japanese-English translation, and "MT" and "mach" are the abbreviations for "machine translation".
ine translation "," natural "
Langue translation system
An example in which "tem" is present is shown. FIG. 4 shows an example of the database for adjusting the length of the translated word in the example shown in FIG. In this database, for each translated word, the length of the character string for generating the translated word is recorded at the same time. FIG. 5 shows an example of an operation of selecting a translated word in the translation system that adjusts the length of a translated sentence, using the examples shown in FIGS. 3 and 4. In this flowchart, the parameter i represents the order of the words to be processed from the beginning among the words arranged in the sentence whose length is to be adjusted.
First, since the translated word is selected for each word of the original sentence, attention is paid to the first word of the original sentence (step S301). When there is only one translated word corresponding to the word, there is no room for selecting the translated word, and therefore the focused word is moved to the next word of the original sentence (steps S303 to S304). When there are a plurality of translated words (step S303), the length of each translated word is adjusted.
Therefore, if the translation result that is output when the translated word length is not adjusted is shorter than the desired output sentence length,
A translated word that is longer than the translated word that is output when the length is not adjusted is selected, and if the translated word is longer, the short word is selected (step S305). Then, the word or phrase to be processed is replaced with the selected word or phrase (step S306). If the length of the sentence output by this selection is compared with the desired length and is within the allowable range (step S307),
Since there is no need to adjust the length anymore, the process ends.
If the difference in length exceeds the allowable range (step S30
7) Then, for further adjustment, move the word of interest to the next word (step S304) and repeat this until the last word of the sentence.

In this way, when the length is adjusted for the translated result, the difference (length difference) between the original sentence and the translated result before the adjustment can be easily found. It is easy to determine whether (summary) should be performed, and the length can be adjusted efficiently.

Further, the lengths may be compared based on the position where the keyword appears in the sentence, and the lengths from the beginning of the sentence to the positions may be compared. For example, by adjusting the summary length so that the positions of the character strings ("fig.1", "Fig. 1", etc.) that refer to the drawings are the same before and after translation, Since the positional relationship between the original sentence and the translated sentence is the same, an easy-to-read translation result can be obtained.
Examples of this are shown in FIGS. 6, 7 and 8. When the original document for translation as shown in FIG. 6 is translated using the conventional translation system and the layout is restored, the page becomes 2 pages as shown in FIG. 7, and the reference point of the drawing is the 2nd page. Has become. On the other hand, when the machine translation device according to the present embodiment performs translation, the layout becomes almost the same as the original sentence as shown in FIG. 8, and a readable translation result is obtained.

Next, a second embodiment of the present invention will be described. FIG. 9 is a block diagram showing a schematic configuration of the machine translation device according to the present embodiment. As shown in FIG. 9, this machine translation device includes an input unit 601 and a display unit 60.
2, a recording unit 603, an editing unit 604, a search unit 605, a document database 606, a summary unit 607, a translation unit 608, and a control unit 609.

The input unit 601 is for inputting data, a command for starting search, etc., and is usually a keyboard or the like. The display unit 602 is for displaying the data input from the input unit 601, the translation result, and the like, and a bitmap display or the like is used. The recording unit 603 is for recording the data input from the input unit 601, the translation result, and the like.

The editing unit 604 performs an editing process on the data input from the input unit 601 and the data recorded in the recording unit. The search unit 605 searches the document database 606 for data that matches the specified conditions. The summarizing unit 607 summarizes the data retrieved by the retrieval unit 605.

The translation unit 608 translates the data summarized by the summarization unit 607 and, if necessary, translates all the data retrieved by the retrieval unit 605. The control unit 609
It controls each of these parts and transfers data. Figure 10
It is a flow chart which shows operation of control part 609 of a machine translation device concerning this example.

First, the control unit 609 starts its operation when a search start command is input from the input unit 601, transfers the search condition A to the search unit 605, and obtains the search result B (step S701). If the search result B is a plurality of documents, it is necessary to confirm the description content of each document in order to select a required document from the documents, so which document should be confirmed from the operator. The document is selected, and the document is designated as C (steps S702 and 703). On the other hand, when the search result B is a single document, it is necessary to confirm whether the search result B is a desired document or not from the description content, and this is designated as C (step S704).

Next, the document C is transferred to the summarizing unit 607, and the summarizing result D is obtained. It is confirmed whether or not the summary result D is written in the mother language of the operator who has been registered in advance (step S706). If it is not the mother tongue, the summary result D is transferred to the translation unit 605 to obtain the translation result E ( Step S
707). In the case of the mother tongue, the summary result D is directly set to E (step S708). Then, E is presented to the operator to end the operation (step S709).

Here, FIG. 11 shows an example of the screen display in this embodiment. In FIG. 11, the uppermost part is a search condition input area, and a search request is made with the keyword “computer” here. In the middle of FIG. 11, a list of titles of documents searched by the above search conditions is displayed. It is instructed to confirm the contents of the second document. Further, although the contents are presented in the lower part, since the translated and translated result is displayed, the user can read only the abstract of this document in the native language of Japanese.

It should be noted that instead of the above configuration, it is also possible to employ a configuration in which a summary is made after translation. However, when adopting a configuration that performs post-summary translation, there is an advantage that the processing time can be shortened because labor for translating parts other than the gist is omitted. Next, a third embodiment of the present invention will be described.

FIG. 12 is a block diagram showing a schematic configuration of a machine translation device according to this embodiment. As shown in FIG. 12, this machine translation device includes an input unit 901, a display unit 902, a recording unit 903, an editing unit 904, and a search unit 90.
5, document database 906, document structure analysis unit 907,
It is configured to have a translation unit 908 and a control unit 909.

The input unit 901 is for inputting data, a command for starting search, etc., and is usually a keyboard or the like. The display unit 902 is for displaying the data input from the input unit 901, the translation result, and the like, and a bitmap display or the like is used.
The recording unit 903 is for recording the data input from the input unit 901, the translation result, and the like.

The editing unit 904 performs an editing process on the data input from the input unit 901 and the data recorded in the recording unit 903. The search unit 905 searches the document database 906 for data that matches the instructed condition.

The document structure analysis unit 907 is for analyzing the structure of the document data retrieved by the retrieval unit 905, and is, for example, the document “A Discosure Struct”.
ure Analyzer for Japanese
Text "(K. Sumita, Int. Con
f. on Fifth Generation co
The structure of the document is obtained by the method described in Mputer '92).

The translation unit 908 translates the data of the portion which is analyzed by the document structure analysis unit 907 and is determined to be important, and translates all the data retrieved by the retrieval unit 905 when necessary. The control unit 909 controls each of these units and transfers data. FIG. 13 is a flowchart showing the operation of the control unit 909 of the machine translation device according to this embodiment.

First, the control unit 909 starts operation when a search start command is input from the input unit 901, transfers the search condition A to the search unit 905, and obtains a search result B (step S1001). If the search result B is a plurality of documents, it is necessary to confirm the description content of each document in order to select a required document from the documents, so which document should be confirmed from the operator. The document is selected, and the document is designated as C (steps S1002 and 1003). on the other hand,
If the search result B is a single document, it is necessary to confirm whether the search result B is a desired document or not from the description content, and this is designated as C (step S1004).

Next, the document C is transferred to the document structure analysis unit 907, and the structure analysis result D is obtained. And the structural analysis result D
A portion judged to be more important in structure is extracted and set as E (step S1006). It is confirmed whether E is described in the native language of the operator registered in advance (step S1007). If it is not the native language, E
Is transferred to the translation section, and the translation result F is obtained (step S10).
08). In the case of the mother tongue, E is directly set to F (step S1009). Then, F is presented to the operator to end the operation (step S1010).

Here, in FIG. 14, the document structure analysis unit 907 is shown.
An example of the result of the structural analysis by is shown. In FIG. 14, the part enclosed by "<" and ">" is a part showing the role of the analysis result, and the sentence which plays the role follows. Whether or not it is structurally important in step S1006 is determined based on, for example, the conclusion part of the sentence or the title. For example, only the conclusion part can be extracted, translated, and presented to the operator. it can. This criterion may be changed by an operator's instruction.

In the above embodiment, Japanese is described as the native language, but other languages may be used. Thus, the present invention can be implemented with various modifications without departing from the scope of the invention.

[0041]

According to the machine translation method of the present invention (Claim 1), since the translation result and the original sentence have the same length, the layout information of the original sentence and the like can be easily reflected in the translated sentence. Further, according to the machine translation method of the present invention (Claim 2), when searching a database other than the native language, since the summary of the search results can be read in the native language, the document search can be performed efficiently and easily. it can.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a machine translation device with a summary function according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing a flow of operation of the machine translation device of FIG.

FIG. 3 is a diagram showing an example in which there are multiple translated words corresponding to original words in Japanese-English translation.

FIG. 4 is a diagram showing an example of a database for adjusting the length of a translated word.

FIG. 5 is a flowchart showing an example of an operation of selecting a translated word.

FIG. 6 is a diagram showing an example of an original sentence to which the embodiment is applied.

FIG. 7 is a diagram showing an example of a translation result by a conventional machine translation method.

FIG. 8 is a diagram showing an example of a translation result according to the same embodiment.

FIG. 9 is a block diagram showing a schematic configuration of a machine translation device with a search / summary function according to a second embodiment of the present invention.

10 is a flowchart showing a flow of operation of the machine translation device of FIG.

FIG. 11 is a diagram showing a screen display example according to the same embodiment.

FIG. 12 is a block diagram showing a schematic configuration of a machine translation device with a search / structure analysis function according to a second embodiment of the present invention.

FIG. 13 is a flowchart showing a flow of operation of the machine translation device of FIG.

FIG. 14 is a diagram showing an example of a structural analysis result according to the example.

[Explanation of symbols]

101 ... Input unit 102 ...
Display unit 103 ... Recording unit 104 ...
Editing unit 105 ... Translation unit 106 ...
Length comparison unit 107 ... Summary unit 108 ...
Control unit 601 ... Input unit 602 ...
Display unit 603 ... Recording unit 604 ...
Editing unit 605 ... Search unit 606 ...
Document database 607 ... Summary section 608 ...
Translation unit 609 ... Control unit 901 ...
Input unit 902 ... Display unit 903 ...
Recording unit 904 ... Editing unit 905 ...
Search unit 906 ... Document database 907 ...
Document structure analysis unit 908 ... Translation unit 909 ...
Control unit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Seiji Miike 1 Komukai Toshiba-cho, Sachi-ku, Kawasaki-shi, Kanagawa Incorporated Toshiba Research and Development Center

Claims (2)

[Claims] 1. A translation step of translating a sentence in a first language into a sentence in a second language, and comparing the length of the sentence in the first language before translation with the length of the sentence in the second language as a result of the translation. Length comparison step, and when the sentence of the second language which is the translation result of the first language is longer than the sentence of the original first language, the sentence of the second language is almost the same as the sentence of the first language. And a summarizing step for summarizing to have the same length. 2. A retrieval step for retrieving a document from a document database described in a first language, a summary step for summarizing sentences retrieved by this retrieval step, and a second document for the document summarized by this summary step. A machine translation method comprising a translation step of translating into a language.