Translation apparatus and method

1. A translation apparatus for translating article information in a first language including an article body and a related headline into a second language, comprising:

a decision unit configured to discriminately identify the article body and the headline;

a phrase alignment processing unit configured to respectively extract a noun phrase from the headline and a noun phrase candidate from a head sentence of the article body, to count coincident words between the noun phrase and the noun phrase candidate, and to output the noun phrase and the noun phrase candidate as a phrase alignment result if a coincidence ratio based on the counted value is above a threshold; and

a translation unit configured to translate the article body and the headline into the second language based on the phrase alignment result.

2. The translation apparatus according to claim 1 ,

wherein said decision unit obtains a uniform resource locator of a Web page of the article information, determines whether the Web page is registered as a news site based on the uniform resource locator, and discriminates the article body and the headline based on a decision algorithm corresponding to the news site if the Web page is registered as the news site.

3. The translation apparatus according to claim 2 ,

wherein said decision unit discriminates the article body and the headline based on a decision algorithm corresponding to a general news page if the Web page is not registered as the news site.

4. The translation apparatus according to claim 1 ,

wherein the coincidence ratio based on the counted value is a ratio of a number of the coincident words to a number of words of the noun phrase candidate.

5. The translation apparatus according to claim 1 ,

wherein said phrase alignment processing unit extracts a noun phrase from the end of the headline, extracts a predetermined pattern including a noun phrase from the head sentence of the article body, and identifies the noun phrase of the headline as an information source part of the article information if the noun phrase of the headline and the noun phrase of the predetermined pattern are included in the phrase alignment result.

6. The translation apparatus according to claim 1 ,

wherein said translation unit translates the headline of the first language into the second language by referring to a special rule for translating the headline.

7. A translation method for translating article information in a first language including an article body and a related headline into a second language, comprising:

discriminately identifying the article body and the headline;

respectively extracting a noun phrase from the headline and a noun phrase candidate from a head sentence of the article body;

counting coincident words between the noun phrase and the noun phrase candidate;

outputting the noun phrase and the noun phrase candidate as a phrase alignment result if a coincidence ratio based on the counted value is above a threshold; and

translating the article body and the headline into the second language based on the phrase alignment result.

8. The translation method according to claim 7 , further comprising:

obtaining a uniform resource locator of a Web page of the article information;

determining whether the Web page is registered as a news site based on the uniform resource locator; and

discriminating the article body and the headline based on a decision algorithm corresponding to the news site if the Web page is registered as the news site.

9. The translation method according to claim 8 , further comprising:

discriminating the article body and the headline based on a decision algorithm corresponding to a general news page if the Web page is not registered as the news site.

10. The translation method according to claim 7 ,

wherein the coincidence ratio based on the counted value is a ratio of a number of the coincident words to a number of words of the noun phrase candidate.

11. The translation method according to claim 7 , further comprising:

extracting a noun phrase from the end of the headline;

extracting a predetermined pattern including a noun phrase from the head sentence of the article body; and

identifying the noun phrase of the headline as an information source part of the article information if the noun phrase of the headline and the noun phrase of the predetermined pattern are included in the phrase alignment result.

12. The translation method according to claim 7 , further comprising

translating the headline of the first language into the second language by referring to a special rule for translating the headline.

13. A computer program product, comprising a memory code embodied in said memory for causing a computer to translate article information in a first language including an article body and a related headline into a second language, said computer readable program code having:

a first program code to discriminately identify the article body and the headline;

a second program code to respectively extract a noun phrase from the headline, and a noun phrase candidate from a head sentence of the article body;

a third program code to count coincident words between the noun phrase and the noun phrase candidate;

a fourth program code to output the noun phrase and the noun phrase candidate as a phrase alignment result if a coincidence ratio based on the counted value is above a threshold; and

a fifth program code to translate the article body and the headline into the second language based on the phrase alignment result.