Identifying related terms in different languages

• US 8,798,988 B1
• Filed: 10/23/2007
• Issued: 08/05/2014
• Est. Priority Date: 10/24/2006
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Upstill, Trystan G., Baker, Steven D.
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): BAKER, MATTHEW H
• Application Number:
  US11/877,407
• Time in Patent Office: 2,478 Days
• Field of Search: 704 1- 10
• US Class Current: 704/9
• CPC Class Codes: G06F 16/90324 using system suggestions
• CPC Class Codes: G06F 40/247 Thesauruses; Synonyms
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation

Abstract

One embodiment of the present invention provides a system that uses a known set of related terms to identify additional related terms in different languages. During operation, the system receives two terms that are known to be related. The system translates these two non-synonym terms into a given language, and then adds the translations of the two related terms to a list of known related pairs.

Claims

1. A computer-implemented method comprising:

receiving a first collection of terms that are designated as cohyponyms, and that are in a first natural language;

generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language;

receiving a candidate pair of terms in the second natural language;

determining, by one or more computers, that both candidate terms of the received pair are present in the second collection of terms; and

determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair, based at least on determining that both candidate terms of the pair are present in the second collection of terms.

2. The method of claim 1, wherein first collection comprise one or more pairs of terms, and wherein each of the one or more pairs of terms contains a first paired term that is a cohyponyms of a second paired term.

3. The method of claim 2, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises generating the second collection of terms by translating each term of the one or more pairs of terms into the second natural language.

4. The method of claim 3, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises:

generating one or more translations or word forms for each term of the one or more pairs of terms;

pairing each of the one or more translations or word forms for the first term of each pair with each of the translations and/or or word forms for the second term of each pair; and

generating the second collection of terms based on the pairing.

5. The method of claim 3, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises:

determining one or more variations for each term of the one or more pairs of terms, wherein a variation of a term is a synonym or normalization of the term that is characterized as a high confidence synonym or normalization, respectively;

generating one or more translations for each variation; and

generating the second collection of terms based on at least the generated one or more translations.

6. The method of claim 2, wherein determining that both candidate terms of the pair are present in the second collection of terms comprises determining that the candidate pair matches a pair of terms included in the second collection of terms.

7. The method of claim 6, wherein determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair comprises:

altering a confidence value associated with the candidate pair of terms to indicate a lower level of confidence that the terms of the pair are synonyms, and

determining to not revise the query to include the second candidate term of the pair based on the altered confidence value.

22. The method of claim 1, wherein the first candidate term is included in the query from a user of an online query system, and the second candidate term is provided by the online query system.

23. The method of claim 1, wherein determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair further comprises determining to not revise a query to include the second candidate term based on a threshold associated with the candidate pair of terms.

8. A non-transitory computer-readable storage medium storing instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

receiving a first collection of terms that are designated as cohyponyms, and that are in a first natural language;

generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language;

receiving a candidate pair of terms in the second natural language;

determining, by one or more computers, that both candidate terms of the received pair are present in the second collection of terms; and

determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair, based at least on determining that both candidate terms of the pair are present in the second collection of terms.

9. The computer-readable storage medium of claim 8, wherein the first collection comprise one or more pairs of terms, and wherein each of the one or more pairs of terms contains a first paired term that is a cohyponyms of a second paired term.

10. The computer-readable storage medium of claim 9, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises generating the second collection of terms by translating each term of the one or more pairs of terms into the second natural language.

11. The computer-readable storage medium of claim 10, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises:

generating one or more translations or word forms for each term of the one or more pairs of terms;

pairing each of the one or more translations or word forms for the first term of each pair with each of the translations or word forms for the second term of each pair; and

generating the second collection of terms based on the pairing.

12. The computer-readable storage medium of claim 10, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises:

determining one or more variations for each term of the one or more pairs of terms, wherein a variation of a term is a synonym or normalization of the term that is characterized as a high confidence synonym or normalization, respectively;

generating one or more translations for each variation; and

generating the second collection of terms based on at least the generated one or more translations.

13. The computer-readable storage medium of claim 9, wherein determining that both candidate terms of the pair are present in the second collection of terms comprises determining that the candidate pair matches a pair of terms included in the second collection of terms.

14. The computer-readable storage medium of claim 13, wherein determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair comprises:

altering a confidence value associated with the candidate pair of terms to indicate a lower level of confidence that the terms of the pair are synonyms, and

determining to not revise the query to include the second candidate term of the pair based on the altered confidence value.

24. The computer-readable storage medium of claim 8, wherein the first candidate term is included in the query from a user of an online query system, and the second candidate term is provided by the online query system.

25. The computer-readable storage medium of claim 8, wherein determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair further comprises determining to not revise a query to include the second candidate term based on a threshold associated with the candidate pair of terms.

15. A computing system comprising:

one or more computers; and

a non-transitory computer-readable storage medium coupled to the one or more computers having instructions stored thereon which, when executed by the one or more computers, cause the one or more computers to perform operations comprising;

receiving a first collection of terms that are designated as cohyponyms, and that are in a first natural language;

generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language;

receiving a candidate pair of terms in the second natural language;

determining, by one or more computers, that both candidate terms of the received pair are present in the second collection of terms; and

determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair, based at least on determining that both candidate terms of the pair are present in the second collection of terms.

16. The computing system of claim 15, wherein the first collection comprise one or more pairs of terms, and wherein each of the one or more pairs of terms contains a first paired term that is a cohyponyms of a second paired term.

17. The computing system of claim 16, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises generating the second collection of terms by translating each term of the one or more pairs of terms into the second natural language.

18. The computing system of claim 17, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises:

generating one or more translations or word forms for each term of the one or more pairs of terms;

pairing each of the one or more translations or word forms for the first term of each pair with each of the translations or word forms for the second term of each pair; and

generating the second collection of terms based on the pairing.

19. The computing system of claim 17, wherein generating a second collection of terms by translating the terms of the first collection that are designated as cohyponyms from the first natural language to a second natural language comprises:

determining one or more variations for each term of the one or more pairs of terms, wherein a variation of a term is a synonym or normalization of the term that is characterized as a high confidence synonym or normalization, respectively;

generating one or more translations for each variation; and

generating the second collection of terms based on at least the generated one or more translations.

20. The computing system of claim 16, wherein determining that both candidate terms of the pair are present in the second collection of terms comprises determining that the candidate pair matches a pair of terms included in the second collection of terms.

21. The computing system of claim 20, wherein determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair comprises:

altering a confidence value associated with the candidate pair of terms to indicate a lower level of confidence that the terms of the pair are synonyms, and

determining to not revise the query to include the second candidate term of the pair based on the altered confidence value.

26. The computing system of claim 15, wherein the first candidate term is included in the query from a user of an online query system, and the second candidate term is provided by the online query system.

27. The computing system of claim 15, wherein determining to not revise a query that includes a first candidate term of the pair, to include a second candidate term of the pair further comprises determining to not revise a query to include the second candidate term based on a threshold associated with the candidate pair of terms.