Word phrase translation using a phrase index

• US 6,473,729 B1
• Filed: 12/20/1999
• Issued: 10/29/2002
• Est. Priority Date: 12/20/1999
• Status: Expired due to Term

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Gastaldo, Michel, Grefenstette, Gregory
• Primary Examiner(s): EDOUARD, PATRICK NESTOR
• Application Number:
  US09/466,823
• Time in Patent Office: 1,044 Days
• Field of Search: 704/2-7, 704/277, 704/10, 707/536, 707/100, 707/532, 434/156, 434/157
• US Class Current: 704/4
• CPC Class Codes: G06F 40/45 Example-based machine translation; Alignment

Abstract

A system and method are provided for translating an input text from a natural source language to a natural target language. The system stores a database that contains a plurality of pairs of text fragments with each pair including a text fragment in the source language and a corresponding text fragment in the target language. Each text fragment contains at least one word phrase and represents a primary grammatical unit such as a sentence or a clause. For translating a word phrase, the database is queried using a phrase index of the database, where the phrase index indexes text fragments by word phrases. Word phrases are noun phrases or word phrases. Alternatively, word phrases are predicates involving at least one verb and one noun or adjective used as a noun. The system further comprises a phrase extractor for extracting a word phrase from a text fragment of an input text.

Claims

1. A method for translating a word phrase from a first natural language to a second natural language, said word phrase being a group of two or more associated words, the method comprising the steps of:

inputting a text written in the first language;

extracting said word phrase from said text;

querying a database for said extracted word phrase using a phrase index of said database;

said phrase index indexing text fragments by word phrases;

said text fragments representing a primary grammatical unit including at least one clause;

the database containing pairs of text fragments, each pair including a text fragment of the first language and a corresponding text fragment of the second language; and

obtaining a translation of said extracted word phrase based on one of the pairs of text fragments revealed during the step of querying the database.

2. The method of claim 1, wherein said word phrase is a noun phrase.

3. The method of claim 1, wherein said word phrase is a verb phrase.

4. The method of claim 1, wherein said word phrase is a predicate involving at least one verb and one noun or adjective used as a noun.

5. The method of claim 1, wherein the step of obtaining a translation includes the step of presenting to a user the revealed pairs of text fragments.

6. The method of claim 1, wherein the step of obtaining a translation includes the step of presenting to a user proposed translations of the extracted word phrase based on the revealed pairs of text fragments.

7. The method of claim 5, wherein the step of obtaining a translation further comprises the step of approving one of the revealed translations by the user and using the approved translation as translation of the extracted word phrase.

8. The method of claim 1, wherein said word phrase is extracted from a text fragment of said input text, and the step of obtaining a translation includes the step of sorting the revealed pairs of text fragments according to the number of word phrases common with said text fragment of said input text.

9. The method of claim 1, wherein the step of extracting a word phrase includes the step of selecting a word of said input text and determining the word phrase comprising said selected word.

10. The method of claim 1, wherein said primary grammatical units are sentences.

11. The method of claim 1, wherein the step of querying the database for the extracted phrase includes the step of querying the database for all word phrases partly matching the extracted phrase.

12. The method of claim 1, wherein the step of obtaining a translation includes querying a terminology base.

13. A computer-readable storage medium storing instructions for translating a word phrase from a first natural language to a second natural language by performing the steps of:

inputting a text written in the first language;

extracting said word phrase from said text, said word phrase being a group of two or more associated words;

querying a database for said extracted word phrase using a phrase index of said database;

said phrase index indexing text fragments by word phrases;

said text fragments representing a primary grammatical unit including at least one clause;

the database containing pairs of text fragments, each pair including a text fragment of the first language and a corresponding text fragment of the second language; and

obtaining a translation of said extracted word phrase based on one of the pairs of text fragments revealed during the step of querying the database.

14. A system for translating an input text from a natural source language to a natural target language, the system comprising:

storage means for storing a database containing a plurality of pairs of text fragments;

said text fragments representing a primary grammatical unit including at least one clause;

each pair including a text fragment in the source language and a corresponding text fragment in the target language, each text fragment containing at least one word phrase, said word phrase being a group of two or more associated words;

a phrase extractor for extracting a word phrase from a text fragment of said input text;

database retrieval means for retrieving, from said database, pairs of text fragments that contain the extracted word phrase, using a phrase index of said database, said phrase index indexing text fragments by word phrases; and

user interface means for allowing a user to select one of said retrieved pairs of text fragments to obtain a translation of the extracted word phrase.

15. The system of claim 14, wherein said word phrase is a noun phrase.

16. The system of claim 14, wherein said word phrase is a verb phrase.

17. The system of claim 14, wherein said word phrase is a predicate involving at least one verb and one noun or adjective used as a noun.

18. The system of claim 14, wherein said user interface means is arranged for presenting to a user the retrieved pairs of text fragments.

19. The system of claim 14, wherein said user interface means is arranged for presenting to a user proposed translations of the determined word phrase into the target language based on the retrieved pairs of text fragments.

20. The system of claim 14, wherein said primary grammatical units are sentences.

21. The system of claim 14, wherein said database retrieval means is arranged for retrieving from the database pairs of text fragments containing word phrases partially matching the extracted word phrase.

22. The system of claim 14, further comprising a terminology base to which said database retrieval means has access.

23. A method for generating a text fragment database for use in translating a word phrase from a first natural language into a second natural language, said word phrase being a group of two or more associated words, the method comprising the steps of:

inputting a first document containing a text written in the first language;

inputting a second document containing said text written in the second language;

aligning corresponding text fragments of the first and second documents;

said text fragments representing a primary grammatical unit including at least one clause;

extracting word phrases from the text fragments of the first document; and

generating index information on the extracted word phrases and the aligned text fragments holding the word phrases, to generate a phrase index indexing text fragments by word phrases.

24. The method of claim 23, wherein said word phrases are noun phrases.

25. The method of claim 23, wherein said word phrases are verb phrases.

26. The method of claim 23, wherein said word phrases are predicates involving at least one verb and one noun or adjective used as a noun.

27. The method of claim 23, wherein said primary grammatical units are sentences.

28. The method of claim 23, wherein the steps of inputting, aligning and extracting are repeated for a plurality of paired documents.

29. The method of claim 23, wherein the contents of said plurality of paired documents are of the same field.