JPH1021222A - Machine translation method and terminology dictionary selecting method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically set environment according to the contents of a document by setting a prescribed condition to be the base of translating processing according to document identification information specifying a document used at the time of obtaining a document to be a translating object. SOLUTION: A control part 2 inputs identification information of a document to translate from an inspecting device 200 by controlling a communication equipment or using communication between processors. Continually the part 2 sends obtained document identification information to a communication part 3 to obtain the content of the document specified by identification information. Next, the part 2 sends document identification information to a field judging part 7 to decide the special field of the obtained document. In addition the part 7 sends an instruction of selecting a terminology dictionary corresponding to the decided special field to a translation part 4. Next, the obtained document is sent to the translation part 4 to execute translation. At the time a translation part 4 translates by using the selected technical word. At last, a translating result is sent to the inspecting device 200 for internet through communication between processors.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a machine translation method for translating a natural language using an electronic computer and a technical term dictionary selecting method for selecting a technical term dictionary used for translation.

[0002]

2. Description of the Related Art In recent years, it has become easy to access overseas information through the Internet. Accordingly, users who translate documents written in English into Japanese using a machine translation system are increasing.

[0003] Since current machine translation systems do not translate 100% of all documents correctly, it is necessary to adjust various parameters of the system so that the translation can be made closer to the translation result desired by the user. A typical example of such a parameter is the selection of a translation dictionary.

At present, many types of machine translation systems can use the following three types of translation dictionaries. One is a “general dictionary” that stores words that are widely used in any specialized field. The other is a “technical term dictionary” that stores terms in a specific technical field such as the information processing field and the chemical field. Still another is a “user dictionary” for registering words and usages unique to the user.

When a machine translation system is used to translate a large number of documents such as manuals in a short time as in the past, the documents translated by the user often relate to a specific technical field. Regarding translation dictionaries,
Once they were set, they were often usable as they were, so there was no need to set them again for each translation. However, when accessing various information published by universities and companies through the Internet, it is not always the case that only documents related to a specific technical field are accessed. Therefore, the user needs to set an appropriate translation dictionary each time to obtain an appropriate translation result. In particular, if there are multiple technical term dictionaries according to the specialty field, the translation result greatly depends on which technical term dictionary is used, and it is necessary to set an appropriate technical term dictionary each time, which is complicated. . Further, not only the setting of the translation dictionary but also the settings related to other translation processes may be switched depending on the specialized field, which is complicated.

[0006]

As described above, in the conventional machine translation system, the user must switch (set a translation environment) various settings that affect the translation result in accordance with the document to be translated. Had to be complicated.

[0007] The present invention has been made in consideration of the above circumstances, and has as its object to provide a machine translation method capable of automatically setting an environment according to the contents of a document. More specifically, an object of the present invention is to provide a technical term dictionary selection method capable of selecting a technical term dictionary used for translation according to the content of a document.

[0008]

Means for Solving the Problems The present invention (claim 1) provides:
A communication device accesses a document in a first language stored on another computer via a network, performs a translation process on the obtained document in the first language based on predetermined conditions, and executes a translation process on the second language. A machine translation method for generating a document according to claim 1, wherein prior to execution of the translation process, the translation process is performed based on document identification information for identifying the document in the first language used at the time of the access. The above-described predetermined condition is set.

A machine translation method according to the present invention (claim 2) receives document identification information for specifying a document to be obtained, obtains the document via a network using the document identification information, A parameter for adjusting a translation result is set based on the document identification information, and the obtained document is translated into a predetermined language based on the set parameter.

[0010] The present invention (claim 3) is based on claim 1 or 2
In the invention described in the above item, the document identification information includes elements of a document format (for example, a protocol name), a host name, a directory name, and a file name, and at least one of the host name, the directory name, and the file name. It is characterized in that the predetermined condition is set based on one element.

The present invention (Claim 4) provides Claims 1 to 3
In the invention according to any one of the above, the predetermined condition indicates a technical term dictionary used at the time of translation.

The present invention (Claim 5) provides Claims 1 to 3
In the invention according to any one of the first to third aspects, the predetermined condition indicates a style to be used in a translated sentence (for example, “mass tone” and “mass tone”).

The present invention (claim 6) provides a technical term used when translating a document in a first language obtained through a network into a second language using document identification information for specifying a document to be obtained. A technical term dictionary selecting method for selecting a dictionary, wherein identification data consisting of all or a part of the document identification information and field data indicating a technical term dictionary to be selected are registered in association with field determination data. It is characterized in that the identification data is compared with the document identification information, and a terminology dictionary to be selected is determined according to the result of the comparison.

According to the present invention (claim 7), in the invention according to claim 6, the document identification information includes elements of a document format (for example, a protocol name), a host name, a directory name, and a file name. And a terminology dictionary to be selected is determined based on at least one of a host name, a directory name, and a file name.

According to the present invention (claim 8), in the invention according to claim 7, the field determination data is a set of a host name, a host name and a directory name, or a set of a host name, a directory name and a file name. , A first field determination data registered in association with a technical term dictionary to be selected, a second field in which a part or all of the directory name is registered in association with the technical term dictionary to be selected In the comparison, a host name, a set of a host name and a directory name of the document identification information, or a set of a host name, a directory name and a file name of the document identification information, and any one of the sets belonging to the first field determination data. If the identification data matches, the technical term dictionary to be selected is determined based on the field data corresponding to the longest matching identification data, and the first term dictionary is determined. If the technical term dictionary cannot be determined from the field determination data, and if any of the identification data belonging to the second field determination data is included in the document identification information, the document identification information The terminology dictionary to be selected is determined based on the field data corresponding to the identification data included in (1).

The invention according to each of the above-described methods is also applicable to the description of the apparatus. According to the present invention, a predetermined condition serving as a base of a translation process is set in accordance with document identification information for specifying a document to be used for obtaining a document to be translated (for example, a technical term used in the translation process). Select a dictionary or specify the style to use in the translation)
This eliminates the need for the user to appropriately set the translation environment according to the content of the document to be translated each time. Therefore, a translation result with the same accuracy can be obtained in a shorter time than in the past. As a result, for example, it is possible to efficiently translate an English document obtained through the Internet and obtain information.

Incidentally, the above-mentioned invention is also realized as a storage medium storing a corresponding program. For example, the present invention allows a communication device to access a document in a first language stored on another computer via a network, and performs a translation process on the obtained document in the first language based on predetermined conditions. And generating a document in a second language by performing document translation on the basis of document identification information for identifying the document in the first language used at the time of the access prior to the execution of the translation processing. A computer-readable recording medium storing a program for controlling a computer to set the predetermined condition serving as a basis of the translation processing.

Further, according to the present invention, the document identification information for specifying the document to be obtained is specified, the document is obtained via a network using the document identification information, and the document identification information is specified based on the document identification information. A computer-readable program that stores a program for controlling a computer to translate the obtained document into a predetermined language based on the set parameters, and to adjust a parameter that can adjust the translation result. It is a recording medium.

Further, the present invention provides a technical term dictionary used when translating a document in a first language obtained through a network into a second language using document identification information for specifying a document to be obtained. A program to be selected, wherein identification data in field determination data in which identification data consisting of all or a part of the document identification information and field data indicating a technical term dictionary to be selected are registered in association with each other; This is a computer-readable recording medium storing a program for controlling a computer so as to compare with document identification information and determine a technical term dictionary to be selected according to the result of the comparison.

[0020]

Embodiments of the present invention will be described below with reference to the drawings. In this embodiment, the original language is English,
An English-Japanese machine translator that translates into Japanese will be described as an example.

In the present embodiment, a case where the original text to be translated is obtained via a network will be described as an example. Here, the network is described as the Internet.

FIG. 1 shows a schematic configuration of a machine translation apparatus according to this embodiment. In this embodiment, as shown in FIG. 1, the machine translation device 100 and the Internet browsing device 200 are used.
Connect and use.

The browsing device 200 is configured using a display device such as a CRT or a liquid crystal display device and an input device such as a mouse or a keyboard device, and has a function of receiving input of identification information of a document to be translated from a user, and a function of receiving the input of the document. It has at least a function of displaying the translation result on the entire screen or in a window in the screen.

The machine translation apparatus 100 of this embodiment translates the downloaded document by accessing the Internet by the communication unit 3 based on the identification information of the document to be translated, which is input from the browsing apparatus 200, and translates the translation result. To the Internet browsing apparatus 200.

The Internet browsing apparatus 200 is, for example, a so-called browser, and in this case, can be realized by executing browser software on a computer. The machine translation device 100 can be realized by executing machine translation software on a computer, for example. In this case, the machine translation software may be executed on (1) a computer that executes the browser software described above, or (2) a machine translation software may be executed on a computer different from the computer that executes the browser software. Conceivable. The communication between the machine translation device 100 and the Internet browsing device 200 is the former (1),
This is realized by inter-process communication. On the other hand, the latter (2)
Is realized through a communication line using a communication device (not shown) provided in each computer.

The transmitting device and the communication unit 3 are:
For example, it is composed of an interface device (network connection device) for connecting to a communication line, and software (network connection software) for performing communication using this interface device. Here, it is assumed that the communication unit 3 is provided in a computer equipped with machine translation software. In the case of the above (2), the communication device 3 may be used as a transmission device necessary for the computer on which the machine translation software is installed, or another communication device may be used.

As shown in FIG. 1, the machine translation apparatus 100 of the present embodiment includes a control unit 2, a translation unit 4, a translation dictionary 6, a field determination unit 7, and field determination data 8.
First, the control unit 2 will be described.

The control section 2 mainly performs the following four controls. (A) Input the identification information of the document to be translated from the browsing device 200 by controlling the communication device as described above or by using the inter-process communication.

(B) The document identification information is passed to the communication unit 3,
The user downloads the document from a network and stores the document in a storage device (not shown)
Etc.).

(C) The obtained document (or pointer information indicating the storage location) is passed to the translator 4, and the identification information of the document is passed for selection of a technical term dictionary. Perform translation using a technical term dictionary.

(D) The result of translation by the translation unit 4 is sent to the Internet browsing device 200 by controlling the communication device or using inter-process communication as described above. Next, the translation unit 4, the translation dictionary 6, the field determination unit 7, and the field determination data 8 will be described.

The translation dictionary 6 stores knowledge information for translation. The translation dictionary 6 is configured using a storage device such as a magnetic disk or an optical disk. The knowledge information stored in the translation dictionary 6 includes, for example, the analysis dictionary 6
a, translation dictionary 6b, translation generation dictionary 6c, original grammar 6d,
There is a translated word grammar 6e.

The translation dictionary 6b further includes a general dictionary 6b1,
It is divided into a technical term dictionary 6b2 and a user dictionary 6b3. In addition, a plurality of technical term dictionaries 6b2 are facilitated in accordance with specialty fields. Here, it is assumed that there are dictionaries in four fields of information, chemistry, sports, and movies as specialized fields, and these are referred to as an information field dictionary 6b2a and a chemical field dictionary 6b
2b, sports field dictionary 6b2c, movie field dictionary 6b2
d.

The general dictionary, each technical term dictionary, and the user dictionary have basically the same configuration. For example, a user dictionary and each technical term dictionary are given priority over the general dictionary. Note that the selected terminology dictionary is used.

The field judging section 7 is for estimating a special field based on the document identification information passed from the control section 2.
At this time, the field determination data 8 is used. The field determination data 8 is configured using a storage device such as a magnetic disk or an optical disk, like the translation dictionary 6.

FIG. 2 is a block diagram showing a basic configuration of the translation unit 4. The translation process is described in detail in, for example, Japanese Patent Application Laid-Open No. 6-96115, entitled "Natural Original Language Processing Apparatus and Electronic Dictionary Apparatus", and only the essential points will be described here.

The translation unit 4 includes a morphological analysis unit 4a, an original sentence structure analysis unit 4b, a structure conversion unit 4c, and a translated sentence generation unit 4d. First, the original sentence is subjected to morphological analysis using the analysis dictionary 6a in the morphological analysis unit 4a, and is decomposed into units for processing in the original language.

Next, the original sentence structure analysis unit 4b performs syntax analysis based on the result of the morphological analysis unit 4a and the original grammar 6d to obtain the syntax structure of the input sentence. After that, the structure conversion unit 4c converts the heading of the original sentence into a translated word or converts the structure of the original sentence into a translated sentence using the translation dictionary 6b. In this embodiment, the selected terminology dictionary is used.

Finally, the translation generation unit 4d generates an original language expression as a translation from the structure of the translation using the translation generation dictionary 6c and the translation grammar 6e. The translation result (translated sentence) is stored in a storage device (not shown) (for example, a hard disk device or a RAM).

Many techniques such as morphological analysis, syntax analysis, and structure conversion have been described in the literature, including the above-mentioned literature, and any of these techniques can be used in the present invention.

Next, the processing procedure of the control unit 2 will be described. FIG. 3 is a flowchart showing the flow of the processing of the control unit 2. In step S1, first, identification information of a document to be displayed on the Internet browsing device 200 is input through a communication device (not shown) or inter-process communication.

FIG. 4 shows an example of the expression format of the document identification information. In the Internet, a format called URL is often used as a format for expressing document identification information. As shown in FIG. 4, the format is a document format (eg, protocol name), a host name, a directory name, or the like. Consists of a file name element. Note that, even in a case other than the Internet, the document identification information can be appropriately configured according to each protocol.

Subsequently, in step S2, step S1
Is transmitted to the communication unit 3 to obtain the contents of the document specified by the identification information. Next, in step S3, the document identification information is sent to the field determination unit 7, and the specialized field of the obtained document is determined by the method described later.
In addition, the field determination unit 7 sends an instruction to select a technical term dictionary corresponding to the determined specialty field to the translation unit 4.

Next, in step S4, the obtained document is sent to the translation unit 4 to be translated. At that time, the translator 4
Then, translation is performed using the selected technical term dictionary.
Finally, in step S5, the translation result is sent to the Internet browsing device 200 through a communication device (not shown) or inter-process communication.

The browsing device 200 displays the received translated document. Next, the field-specific determination data will be described. FIG. 5 is a diagram illustrating an example of the content format of the “determination data for each field” and a specific example of the content.

Each element of the field determination data is composed of three columns of "data type", "identification data", and "field name". The “data type” indicates the type of data stored in the “identification data” column, and stores a character “H” or “D”.

When the data type is “H”, the “identification data” column contains (1) host name, (2) host name + directory name, and (3) host name of the document identification information specified by the URL. Stores any one of name + directory name + file name.

When the data type is "D", all or part of the directory part of the field identification information is stored as "identification data". In the “field name” column, a field name corresponding to the field corresponding to “identification data” regardless of the data type is stored.

It should be noted that, even in cases other than the Internet, it is possible to appropriately configure the discrimination data for each field according to each protocol. Next, a processing procedure of the field determination unit 7 will be described.

FIG. 6 is a flowchart showing the flow of the process of the field determination section 7. This process is roughly divided into two processes. The process up to step S110 is a process of searching for data whose "data type" is "H", and the process after step S120 is a process of searching for data whose "data type" is "D". is there.

First, in step S 101, the “document format” portion is removed from the document identification information in the URL format received from the control unit 2, and only the data after the “host name” is substituted for the variable X.

The following steps S102 to S10
In step 9, a search is made for field determination data that best matches the given document identification information. That is, step S1
In 02, 0 is set to each of the variables I and L. Variable I
Is used to refer to the element I of the field determination data. The variable L is for storing the number of matching characters from the beginning of the character string in the document identification information and the field determination data assigned to the variable.

In the check in step S103,
If there is data for field determination that has not been compared, it is checked in step S103b whether the data type is "H". If H, in step S104, the length of the identification data of the element I of the field determination data is substituted for the variable N, and in step S105, the identification data substituted for the variable X in step S101 and the first N characters are used. Compare for a match. If not, the value of the variable I is incremented by 1 in step S109, and the process returns to step S103.

If the first N characters match in step S105, it is checked in next step S106 whether the value of the variable N is larger than the value of the variable L. If the value of N is small, the process proceeds to step S109. If the value of the variable N is large, the value of the variable N is substituted for the variable L in step S107, and the field name of the element I of the field determination data is substituted for the variable T.

As described above, by the processes of steps S102 to S109, the specialized field which best matches the document identification information given from the control unit 2 is searched. Next, step S
At 110, it is checked whether the value of the variable L is 0. If it is 0, it means that the specialized field could not be determined. In that case, the process proceeds to step S120 to determine the field using the data of the data type D.

On the other hand, if L = 0 is not satisfied in step S110, the flow advances to step S111 to send the field name stored in the variable T. The translator 4 selects a technical term dictionary corresponding to the field based on the field name sent.

In step S120, the value of the variable I is reset to 0 in order to search the field determination data from the beginning.
If it is determined in step S121 that the field determination data that has not been compared remains, the process proceeds to step S12.
Step 2 checks whether the data type is "D". If D, it is checked in step S123 whether the identification data substituted in the variable X in step S101 includes the identification data of the element I of the field determination data. If not, the value of the variable I is incremented by 1 in step S124, and the process returns to step S121.

If it is determined in step S123 that the identification data is included, the field name of the element I of the field determination data is substituted for the variable T in step S125. The field name stored in the variable T is sent to the translator 4 and the process is terminated.

In the above procedure, if a field name is not finally selected, (1) use the same terminology dictionary selected last time, (2) set a default, Various methods can be employed, such as using a technical term dictionary or (3) not using a technical term dictionary.

As described above, according to the present embodiment, it is not necessary to designate a specialized field dictionary corresponding to the contents of a document each time a translation is performed as in the prior art. Can be efficiently translated and information can be obtained.

The present invention is not limited to the above embodiment. In the above description, one specialized field dictionary was selected. However, by registering a plurality of fields as necessary in the field of the field name of the field-specific determination data, a plurality of specialized field dictionaries can be selected according to the identification-specific data. Is also possible. When a plurality of specialty dictionaries are selected, priorities for using these specialty dictionaries in translation are registered in the discrimination data for each field, and the translator 4 assigns a plurality of specialty dictionaries in accordance with the registered priority levels. Preferably, a field dictionary is used.

In the present embodiment, the specialized field is determined using the document identification information, and the specialized field dictionary is specified based on the determined field. However, using the knowledge described in the same format,
It is also possible to change various parameters associated with the translation system. For example, it is also possible to estimate the type of document (manual, news, commentary, product description, etc.) and specify the style of the translated text (“Dama-San” and “Dana-Naru”) accordingly. It is. Alternatively, it is also possible to estimate the type of the document and specify the expression form in the translated sentence (for example, the difference between “-do” and “-do” when translating an English imperative form). It is. In addition, the present invention can be applied to various settings related to translation processing.

In this embodiment, the field is determined using only data registered in advance as field determination data. However, by using a display device such as a CRT and an input device such as a keyboard and a mouse, the field can be determined. If it is not possible to make a determination using the data for field determination, ask the user,
It is also possible to add the result as field determination data.

Hereinafter, a description will be given of hardware and software configurations when the machine translation apparatus of the present embodiment is implemented on a computer. The machine translation device of the present embodiment has a hardware configuration as described above,
It can be configured using a CPU, a RAM for storing programs and necessary data, an input / output device, a storage device such as a hard disk device, and a communication device.
The control unit 2, the translation unit 4, and the field determination unit 7 in FIG.
Is a program describing these processing procedures (for example, the procedures in FIGS. 2 and 3), and is realized by a program for controlling a computer. Also, various dictionaries and field determination data are stored in a storage device such as a hard disk device. The obtained original text and translation result are stored in a storage device such as a hard disk device or a RAM.

Then, the above-mentioned program is stored in the RAM and executed by the CPU. The CPU performs operations, input / output devices,
A desired function is realized by controlling a communication device or a storage device.

The above program may be stored in a ROM, but may be realized by installing the program in a general-purpose computer as follows.
That is, first, the program is stored in a computer-readable recording medium (for example, a recording medium such as a floppy disk or a CR-ROM), and the program is read using a disk drive device corresponding to the recording medium. Store in RAM and execute. Alternatively, the program is temporarily installed in a hard disk device or the like, and is stored in the RAM from the hard disk device or the like at the time of execution and executed. If the recording medium storing the program is an IC card, the program can be read using an IC card reader. Alternatively, a program can be received from a predetermined interface device via a network.

As described above, the program for the machine translation device and the program for the browsing device can be realized on different computers or on the same computer.

In the above embodiment, an English-Japanese machine translation apparatus in which the original language is English and the translated word is Japanese has been described as an example. However, the present invention does not limit the combination of the original language and the translated word. The present invention can be applied to various machine translators such as a machine translator that uses German as the original language and Japanese as the translation, or a machine translator that uses Japanese as the original language and English as the translation. The present invention is not limited to the above-described embodiment, and can be implemented with various modifications within the technical scope.

[0069]

According to the present invention, a predetermined condition (for example, a technical term used in translation) is used as a basis for translation processing in accordance with document identification information for identifying a document used for obtaining the document to be translated. Since the dictionary is set, the user does not need to appropriately set the translation environment according to the content of the document to be translated each time as in the related art. Therefore,
It is possible to translate and browse information in a shorter time than before.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of a machine translation system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a basic configuration of a translation unit according to the embodiment;

FIG. 3 is a flowchart showing a flow of processing in a control unit.

FIG. 4 is a diagram showing an example of an expression format of document identification information.

FIG. 5 is a diagram showing a data structure of field determination data and an example of the data;

FIG. 6 is a flowchart showing a flow of processing in a field determination unit.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Input part 2 ... Control part 3 ... Communication part 4 ... Translation part 5 ... Output part 6 ... Translation dictionary 7 ... Field determination part 8 ... Field determination data 100 ... Machine translation apparatus 200 ... Internet browsing apparatus

Claims (8)

[Claims] 1. A document in a first language stored on another computer via a network by a communication device is accessed,
A machine translation method for performing a translation process based on a predetermined condition on the obtained document in the first language to generate a document in a second language, wherein, prior to performing the translation process, A machine translation method, wherein the predetermined condition serving as a basis for the translation processing is set based on document identification information for specifying a document in the first language to be used. 2. Receiving designation of document identification information for specifying a document to be obtained, obtaining the document via a network using the document identification information, and translating a translation result based on the document identification information. A machine translation method comprising: setting adjustable parameters; and translating the obtained document into a predetermined language based on the set parameters. 3. The document identification information includes elements of a document format, a host name, a directory name, and a file name, and is based on at least one of the host name, the directory name, and the file name. The machine translation method according to claim 1, wherein the predetermined condition is set by setting. 4. The system according to claim 1, wherein the predetermined condition indicates a technical term dictionary used for translation.
4. The machine translation method according to any one of items 3 to 3. 5. The machine translation method according to claim 1, wherein the predetermined condition indicates a style to be used in a translated sentence. 6. A technical term dictionary for selecting a technical term dictionary used when translating a document in a first language obtained through a network into a second language using document identification information for specifying a document to be obtained. A selection method, wherein identification data in field determination data registered in association with identification data consisting of all or a part of the document identification information and field data indicating a technical term dictionary to be selected; A technical term dictionary selecting method comprising comparing information with information and determining a technical term dictionary to be selected according to the result of the comparison. 7. The document identification information includes elements of a document format (eg, protocol name), a host name, a directory name, and a file name, and at least one of a host name, a directory name, and a file name. 7. The technical term dictionary selection method according to claim 6, wherein a technical term dictionary to be selected is determined based on the element. 8. The data for discriminating fields includes a host name, a set of a host name and a directory name, or a set of a host name, a directory name and a file name and a terminology dictionary to be selected and registered in association with each other. 1 field determination data, part or all of a directory name, and second field determination data registered in association with a terminology dictionary to be selected. In the comparison, the host of the document identification information is used. If the set of name, host name and directory name, or the set of host name, directory name and file name matches any of the identification data belonging to the first field determination data, the longest matching identification A terminology dictionary to be selected is determined based on field data corresponding to the data,
The technical term dictionary could not be determined by the field judgment data of
In the case where there is identification data belonging to the document identification information among the identification data belonging to the second field determination data, the identification data is selected by the field data corresponding to the identification data included in the document identification information. 8. The technical term dictionary selection method according to claim 7, wherein a technical term dictionary to be determined is determined.