JP5468576B2 - Translation apparatus, translation method, and translation program - Google Patents

Description

  The present invention relates to automatic translation.

  In recent years, automatic translation using a computer has become widespread. The outline of automatic translation is based on the translation dictionary, which is the data for translating the source language into the target language, replacing the words contained in the original text with the corresponding translated words, and according to the syntax and meaning of the analysis of the original text. This is a technology for outputting as translated sentences. In the field of such automatic translation, various ideas for improving the quality of translation have been proposed. For example, when there are a plurality of translation word candidates corresponding to the source language in the target language, the image search result for each candidate is displayed to assist selection of an appropriate translation word reflecting the user's search intention. A device is known (for example, refer to Patent Document 1).

JP 2010-198525 A

National Institute of Information and Communications Technology, "EDR Electronic Dictionary Specification Manual (Chapter Unit Download)", [online], [Search February 25, 2011], Internet <URL: http: //www2.nict .go.jp / r / r312 / EDR / JPN / TG / TG-View.html>

  In the above example, for words included in the translation dictionary, it may be helpful to select a translated word. However, in reality, words that are not included in the translation dictionary, such as nicknames of celebrities, proper nouns on works, various coined words, abbreviations, slang, etc. In many cases, the original sentence includes “unknown word”. Conventionally, such unknown words are simply output as included in the part corresponding to the unknown word in the translated sentence, for example, as Roman letters according to the language to be translated, so that it is not useful for understanding the translation result was there.

  In view of the above problems, an object of the present invention is to assist in understanding the translation result of an unknown word that is not in the translation dictionary.

Based on the above object, a translation apparatus according to one aspect (1) of the present invention includes a translation dictionary storage unit that stores a translation dictionary that is data for translating a translation source language into a translation destination language, A source sentence acquiring means for acquiring a source sentence of a language; a translation means for translating the acquired source sentence into a translated sentence in a translation destination language using the translation dictionary; and an unknown word not included in the translation dictionary from the source sentence An unknown word detection means for detecting, an image search means for acquiring an image search result for the detected unknown word, and an image that is the image search result at a location corresponding to the unknown word in the translated sentence And a result output means for outputting the translated sentence inserted in place of the unknown word .

According to another aspect (6) of the present invention, the above-described aspect is grasped in the category of the method, and a translation dictionary storage means for storing a translation dictionary which is data for translating the translation source language into the translation destination language. A translation method using a translation device having an original sentence acquisition process for acquiring an original sentence in a translation source language, and a translation process for translating the acquired original sentence into a translation sentence in a target language using the translation dictionary; An unknown word detection process for detecting an unknown word not included in the translation dictionary from the original sentence , an image search process for obtaining an image search result for the detected unknown word, and corresponding to the unknown word among the translated sentences The computer executes a result output process of outputting a translated sentence in which the image as the image search result is inserted in place of the unknown word at a place to be processed.

According to another aspect (7) of the present invention, the above aspect is captured in the category of a computer program. In a computer translation program, an original sentence acquisition process for acquiring an original sentence in a source language, and the acquired original sentence A translation process for translating the translation source language into the translation language using the translation dictionary, which is data for translating the translation source language into the translation destination language, and an unknown word that is not included in the translation dictionary from the source sentence A word detection process, an image search process for obtaining an image search result for the detected unknown word, and an image that is the image search result in the portion corresponding to the unknown word in the translated sentence. Instead, a result output means for outputting the inserted translated sentence is executed by a computer.

  In this way, for unknown words that are not in the translation dictionary, by outputting the image search results in addition to the translation results, anyone can understand the contents as auxiliary information that does not depend on the language. It becomes possible to do.

  According to another aspect (2) of the present invention, in any one of the above aspects, the result output means outputs the image that is the image search result as the acquired image search result in the display of the translated sentence. It is characterized by being replaced with a word.

  In this way, by replacing the image of the search result with the unknown word part of the translated sentence and outputting it, it can be naturally grasped as an element included in the context of the translated sentence. It is possible to assist.

  According to another aspect (3) of the present invention, in any one of the above aspects, the image search means transmits a search request using the unknown word as a search word to the image search server of the translation source language. And

  In this way, by causing the image search server in the source language to perform an image search using the unknown word as a keyword, the image search server in the source language should have a lot of information such as a search index for the unknown word. For this reason, it is possible to obtain a more complete image search result for an unknown word and to assist the understanding of the translation result more effectively.

  According to another aspect (4) of the present invention, in any one of the above aspects, the result output unit displays the translated sentence on a screen, and a predetermined part of the displayed translated sentence corresponding to the unknown word is specified. The image search result is output when an operation is input.

  As described above, by displaying the image search result through a predetermined operation such as clicking, an image is not displayed from the beginning in the translated sentence. For this reason, there is no inconvenience that the line feed pitch spreads out at the height of the image and damages the list of translated sentences from the beginning, and the image can be displayed at the timing determined by the user even if the packet communication fee or attention is concerned. Excellent usability can be obtained.

  The translation apparatus according to another aspect (5) of the present invention, in any one of the above aspects, includes a translation dictionary storage unit that stores a translation dictionary that is data for translating the translation source language into the translation destination language. Included in the translation dictionary from the original sentence, the original sentence acquisition means for acquiring the original sentence in the source language, the translation means for translating the acquired original sentence into the first translated sentence in the target language using the translation dictionary An unknown word detecting means for detecting an unknown word, a web search means for obtaining a web search result for the detected unknown word, and a second translation obtained by translating the obtained web search result into the target language And a result output means for outputting a sentence regarding a portion corresponding to the unknown word in the first translated sentence.

  In this way, even unknown words that are not in the translation dictionary can be translated even by those who understand only the target language by outputting the second translated sentence translated from the web search result into the target language in addition to the translation result. It is possible to understand the contents such as the positioning of the unknown word from the translation of the search result of the unknown word in the original language area, and to assist in understanding the translation result.

  It should be noted that categories different from the above-described aspects (methods for apparatuses, programs for methods, etc.) and more specific aspects described below are also included in the present invention. For categories different from the above-described aspects, “means” is appropriately read as “processing” or “step”. Further, the order of execution of processes and steps is not limited to the above, and can be changed as appropriate or can be processed collectively. Furthermore, in the category of methods and programs, “computers” that execute individual processes may be common or may be different for each process.

  According to the present invention, it is possible to assist understanding of translation results even for unknown words that are not in the translation dictionary.

The functional block diagram which shows the structure of embodiment (1st Embodiment) of this invention. The figure which illustrates the information (data) in the embodiment of the present invention. The flowchart which shows the process sequence in embodiment of this invention. The figure which shows the example of a screen display in embodiment of this invention. The figure which shows the example of a screen display in 2nd Embodiment of this invention.

  Next, as an example of the present invention, a mode for carrying out the present invention (referred to as “embodiment”) will be described with reference to the drawings. It should be noted that assumptions common to those already described in the background art and problems are omitted as appropriate.

[1. Constitution〕
The present embodiment relates to a translation apparatus 1 (also referred to as “this apparatus” or “this apparatus 1”) shown in FIG. 1 (configuration diagram). This apparatus 1 performs translation with respect to a terminal T via a communication network N. Although it is an example of a web server that provides a service, it may be realized as stand-alone software or a device. The configuration diagram of FIG. 1 also shows an example of contents such as information related to processing in a broken-line rectangle on the right side.

  First, the apparatus 1 has at least an arithmetic control unit 6 such as a CPU, a storage device 7 such as a main memory and an auxiliary storage device, and a communication network N (for example, the Internet, a mobile phone network, a PHS network, etc.) as a computer configuration. ) Communication means 8 (communication gateway device, communication circuit with mobile phone network or PHS network, wireless LAN adapter, etc.). Further, although not shown, the terminal T is an electronic information device having an input / output unit using a liquid crystal display panel, a touch panel, a push button, etc. in addition to the above-described computer configuration, such as a smartphone, a mobile phone terminal, These are tablet PCs, mobile PCs, stationary PCs, and the like. Although a small number is shown in FIG. 1, there are actually many according to the number of users.

  On the other hand, in the present apparatus 1, elements (10, 20, etc.) such as each means shown in FIG. 1 are realized by a predetermined computer program stored (installed) in the storage device 7 controlling the arithmetic control unit 6. . Of these elements, the information storage means can be realized in the storage device 7 in any format such as various databases (also referred to as “DB”), variables such as files and arrays, various stacks and registers, and system setting values.

  Among such storage means, the translation dictionary storage means 25 is a means for storing a translation dictionary which is data for translating the translation source language into the translation destination language. The content of the translation dictionary may be the same data as in the past, so illustration is omitted, but basically (for example, refer to Non-Patent Document 1), notation of a word (for example, “February”), part of speech, pronunciation, It includes a semantic attribute and a translated word (“February”) that is contrast (parallel translation) data, and the translated word also includes notation, part of speech, pronunciation, semantic attribute, and the like.

  In addition, each means other than the storage means is a processing means for realizing and executing the following information processing functions / actions, but these are functional units arranged for explanation, actual hardware elements and It doesn't matter if it matches the software module.

[2. Action and effect)
[2-1. Overview〕
In the present embodiment configured as described above, as shown in the flowchart of FIG. 3, first, the source text acquisition means 10 acquires the source text (for example, content example ex1 in FIG. 1) of the source language (Japanese). Then (step S1), the translation means 20 translates the acquired original sentence into a translated sentence (for example, content example ex2) in the translation destination language (English) using the translation dictionary (step S2). Regarding the acquisition of the original text, the mode such as typing in the input field and uploading the file is free.

  When the unknown word detection means 30 detects an unknown word that is not included in the translation dictionary from the original text (step S3: “YES”), the image search means 40 acquires an image search result for the detected unknown word ( Step S4). For example, in the example of FIG. 1, if the word “Kimtaku” in the content example ex1 that is the original sentence is not in the translation dictionary, the word “Kimutaku” (content example ex3) must be written in Roman letters in the content example ex2 that is the translated sentence. Absent. In this case, the image search means 40 acquires an image search result based on the detected unknown word (“Kimtaku” when searching in the original Japanese) (step S4), and the result output means 50 is acquired. The image search result is output for the portion corresponding to the unknown word in the translated sentence (step S5, for example, content example ex5). Here, the image search result to be output may be a web page of an image search result including a plurality of images hit by the search, or may be an individual image itself hit.

  Thus, for unknown words that are not in the translation dictionary, by outputting the image search result in addition to the translation result (for example, step S5), the image can be understood by anyone as auxiliary information independent of the language. It is possible to assist in understanding the results.

[2-2. (Search images)
In particular, it is desirable that the image search result acquired by the image search means 40 for an unknown word (step S4) is typically the result of an image search by a so-called Internet search engine (so-called search site, net search). In this case, the image search means 40 provides a search request using an unknown word (character string “Kimtaku” in the above example) as a search word, and provides an image search service in the translation source language (for example, Japanese). For example, the image is transmitted to the image search server 2.

  As described above, by causing the image search server (for example, 2) in the translation source language to perform an image search using the unknown word as a keyword, the image search server in the translation source language has a large amount of information such as a search index for the unknown word. Therefore, it is possible to obtain a more complete image search result for the unknown word and to assist in understanding the translation result more effectively.

  However, “image search” in the image search result is not limited to that by a search engine. For example, a person name matching an unknown word as a key is searched from a person name / image contrast dictionary (for example, FIG. 2), Search results such as acquiring corresponding images can be used, and in this way, images that are determined to be appropriate in advance are stable regardless of whether or not there is an image that is hit by the search engine regardless of the search engine. Can be used. In addition, in the case where there are a plurality of corresponding images, including the case of net search, the criterion for selecting the image to be acquired or the image to be output may be the first image or an image having many similar images. , Including the number of images to be selected.

[2-3. Output mode]
In the example of FIG. 1, the image is displayed from the beginning as the acquired image search result (content example ex5 in FIG. 1). As another output example, the image search result is obtained through a predetermined operation. It may be displayed. In this case, as illustrated in FIG. 4A, when the result output means 50 displays the translated sentence T0 on the window G1 or the like on the screen, for example, the hyperlink L is displayed on the unknown word portion or a nearby alert is issued. Icon M is displayed. Then, the result output means 50 receives a predetermined operation such as a click operation with the mouse cursor C on the hyperlink L or the alert icon M corresponding to the unknown word in the displayed translated sentence T0. In addition, as illustrated in FIG. 4B, the image search result is output to another window G2 or the like.

  At this time, the display of the image search result is not limited to another window G2, but may be a form in which the contents of the window G1 are changed over to a different page, or a window in the window G1 without page transition. G2 may be displayed in a nested manner, or a new window that is independent from window G1 may be popped up. However, in any case, it is not essential to display the image search result in a separate window. The image may appear below or near the translated sentence T0 without explicitly dividing the window or the display area.

  As described above, by displaying the image search result through a predetermined operation such as clicking, an image is not displayed from the beginning in the translated sentence. For this reason, there is no inconvenience that the line feed pitch spreads out at the height of the image and damages the list of translated sentences from the beginning, and the image can be displayed at the timing determined by the user even if the packet communication fee or attention is concerned. Excellent usability can be obtained. In this case, the timing for obtaining the image search result may be searched and output in real time at the time of operation, or the search result obtained in advance may be output after waiting for the operation. But you can.

  Of course, as shown in a content example ex5 in FIG. 1, it is intuitive for the user that the image G, which is the image search result in FIG. It is desirable in terms of having the meaning understood.

  In this way, by replacing the image of the search result with the unknown word part of the translated sentence and outputting it, it can be naturally grasped as an element included in the context of the translated sentence. It is possible to assist.

[3. Second Embodiment]
In the above embodiment (referred to as the first embodiment), the understanding of the unknown word is assisted with an image (for example, the content example ex5 in FIG. 1), but the understanding of the unknown word can also be assisted in the translated language. (Second Embodiment). Regarding the configuration (not shown) of the second embodiment, the translation dictionary storage means, the original sentence acquisition means, and the unknown word detection means may be the same as in the first embodiment, but the translation means and the result output means are as follows. Thus, the contents are different, and the image search means is replaced with the web search means described later.

  That is, in the second embodiment, the translation unit translates the acquired original sentence into the first translated sentence in the translation destination language using the translation dictionary, and the web search unit performs the web search result for the detected unknown word. The result output means outputs a second translated sentence obtained by translating the obtained web search result into the translation target language with respect to a portion corresponding to the unknown word in the first translated sentence.

  For example, FIG. 5 (1) is an example of a screen displaying the first translated sentence T1, and when the hyperlink L of the unknown word portion or the alert icon M nearby is operated with the mouse cursor C, A second translated sentence T2 obtained by translating the unknown word on the web with the original word “Kimtaku” (FIG. 5 (2)) into English as the translation destination language is displayed on the screen (FIG. 5 (3)). In this example, the web search result TX in FIG. 5B is a result of using an unknown word included in the original text as a search word using the original word “Kimtaku”, and is not a screen display target.

  In this way, for unknown words that are not in the translation dictionary, the second translated sentence obtained by translating the web search result into the translation destination language is output in addition to the translation result (for example, FIG. 5 (3)), Even a person who understands only the language can understand the contents such as the positioning of the unknown word from the translation of the search result of the unknown word in the translation source language area, and can assist the understanding of the translation result.

  In addition, the glossary is optimal for outputting as the second translated sentence as described above. In this case, the corresponding URL of the web dictionary may be registered in advance, and a web page corresponding to such a specific URL may be selected and displayed as a web search result, or a glossary search A search result corresponding to an unknown word may be displayed on a specific URL such as a reception page by passing an unknown word as a parameter using an API or the like.

  5 (2) and FIG. 5 (3) exemplify individual web pages called glossaries as web search results. However, a list of web pages hit by the search is translated into the target language. It may be a second translation. Such a list of web pages generally includes a summary (a so-called snippet) regarding the search keyword “Kimtaku”, which is useful for understanding unknown words.

  Also, the second translated sentence T2 shown in FIG. 5 (3) may be translated by the same translation means as the first translated sentence, or the URL of an automatic translation result page by another system, etc. May be used. Further, FIG. 4 includes HTML source text that displays the second translated sentence T2 as shown in FIG. 5 (3), a hyperlink L that transitions to such a second translated sentence T2, a warning icon M, and the like. If a means for outputting the source text of HTML for realizing the display as shown in (1) and providing it to the user is provided, the utilization result can be flexibly expanded by incorporating the translation result according to the present invention into an arbitrary web page. be able to.

[4. Other embodiments]
In addition, the said embodiment is only an illustration and this invention includes what is illustrated below and other embodiment other than that. For example, the configuration diagrams, data diagrams, flowcharts, and the like in the present application are merely examples, and the presence / absence of each element, the order of the arrangement and processing execution, and the specific contents can be changed as appropriate.

  As an example, a configuration in which the individual units shown in FIG. 1 and the like are realized by separate and independent facilities is also common, and depending on the function of the server or the terminal, an external platform or the like may be changed to an API (application program program). The configuration can be flexibly changed, for example, by calling with an interface) or network computing (so-called cloud or the like). Furthermore, each element such as means relating to the present invention may be realized by other information processing mechanisms such as a physical electronic circuit as well as an arithmetic control unit of a computer.

1 Translation device (this device)
2 Image search server 6 Arithmetic control unit 7 Storage device 8 Communication means 10 Original sentence acquisition means 20 Translation means 25 Translation dictionary storage means 30 Unknown word detection means 40 Image search means 50 Result output means C Mouse cursors ex1 to ex5 Content examples G1, G2 Window L Hyperlink M Alert icon T Terminals T0, T1, T2 Translated text T2 Translated text

Claims (6)

A translation dictionary storage means for storing a translation dictionary which is data for translating the source language into the destination language;
An original text acquisition means for acquiring the original text of the source language;
A translation means for translating the acquired original sentence into a translated sentence in a target language using the translation dictionary;
Unknown word detection means for detecting unknown words not included in the translation dictionary from the original text;
Image search means for obtaining an image search result for the detected unknown word;
A result output means for outputting a translation sentence in which the image as a result of the image search is inserted instead of the unknown word at a position corresponding to the unknown word in the translation sentence;
A translation apparatus comprising: The translation apparatus according to claim 1 , wherein the image search means transmits a search request using the unknown word as a search word to the image search server of the translation source language. The result output means displays the translated sentence on a screen, and outputs the image search result when a predetermined operation is input for a portion corresponding to the unknown word in the displayed translated sentence. The translation apparatus according to claim 1 or 2 . A person image storage means for storing the person image and the person name in association with each other;
The translation apparatus according to claim 1, wherein the image search unit searches for an image associated with the unknown word from a person image stored in the person image storage unit. A translation method using a translation device having translation dictionary storage means for storing a translation dictionary which is data for translating a translation source language into a translation destination language,
Source acquisition processing to acquire source language source text,
A translation process for translating the acquired original sentence into a translated sentence in a target language using the translation dictionary;
An unknown word detection process for detecting unknown words not included in the translation dictionary from the original text;
Image search processing for acquiring an image search result for the detected unknown word;
A result output process for outputting a translated sentence in which the image as a result of the image search is inserted in place of the unknown word at a location corresponding to the unknown word in the translated sentence;
A translation method characterized in that a computer executes the above. In computer translation programs,
Source acquisition processing to acquire source language source text,
A translation process for translating the acquired original sentence into a translated sentence in the translation destination language using a translation dictionary which is data for translating the translation source language into the translation destination language;
An unknown word detection process for detecting unknown words not included in the translation dictionary from the original text;
Image search processing for acquiring an image search result for the detected unknown word;
A result output process for outputting a translated sentence in which the image as a result of the image search is inserted in place of the unknown word at a location corresponding to the unknown word in the translated sentence;
A translation program characterized by causing a computer to execute.

Images