Lexicon development via shared translation database
A speech translation system and methods for cross-lingual communication that enable users to improve and customize content and usage of the system and easily. The methods include, in response to receiving an utterance including a first term associated with a field, translating the utterance into a second language. In response to receiving an indication to add the first term associated with the field to a first recognition lexicon, adding the first term associated with the field and the determined translation to a first machine translation module and to a shared database for a community associated with the field of the first term associated with the field, wherein the first term associated with the field added to the shared database is accessible by the community.
1. A method comprising:
receiving, from a user, an utterance in a first language, the utterance including a term, the utterance associated with an indication from the user to add the term to a user's individual lexicon of a speech recognition system;
determining a word class for the term;
determining a first value of an intra-class probability for the term, the intra-class probability reflecting a relevance of the term specific to the user relative to other terms in the user's individual lexicon that belong to the word class;
increasing the intra-class probability of the term from the first value to a second value based on the indication from the user to add the term to the user's individual lexicon, the increase in the intra-class probability representing the term being more likely to be used by the user due to the indication from the user to add the term to the user's individual lexicon;
adding the term, the word class, and the intra-class probability to the lexicon of the user.
2. The method of claim 1 , wherein the intra-class probability is determined based on when the utterance was received from the user.
3. The method of claim 1 , wherein the first value of the intra-class probability is determined based on reference information about the first term.
4. The method of claim 3 , wherein the first term is a name of a city and the reference information is a population of the city.
5. The method of claim 3 , wherein the reference information is one or more references to the first term in media.
6. The method of claim 1 , wherein the first value of the intra-class probability is determined based on one or more activities of the user.
7. The method of claim 6 , wherein the one or more activities includes a current location of the user.
8. The method of claim 7 , wherein the first term is a place and the first value of the intra-class probability is determined based on a distance between the place and the current location of the user.
9. The method of claim 7 , wherein the first value of the intra-class probability is determined based on a correlation between the first term and the current location of the user.
10. The method of claim 6 , wherein the one or more activities includes past use of the first term by the user.
11. The method of claim 1 , wherein the second value of the intra-class probability decays over time.
12. The method of claim 1 , further comprising:
updating the second value of the intra-class probability responsive to new information.
13. The method of claim 12 , wherein the new information is an updated location of the user.
14. The method of claim 1 , further comprising:
translating the first term from the first language to a second language,
wherein the translated first term is also added to the lexicon of the user.
15. A non-transitory computer-readable medium comprising instructions that when executed by a processor cause the processor to perform steps comprising:
receiving, from a user, an utterance in a first language, the utterance including a term, the utterance associated with an indication from the user to add the term to a user's individual lexicon of a speech recognition system;
determining a word class for the term;
determining a first value of an intra-class probability for the term, the intra-class probability reflecting a relevance of the term specific to the user relative to other terms in the user's individual lexicon that belong to the word class;
increasing the intra-class probability of the term from the first value to a second value based on the indication from the user to add the term to the user's individual lexicon, the increase in the intra-class probability representing the term being more likely to be used by the user due to the indication from the user to add the term to the user's individual lexicon;
adding the term, the word class, and the intra-class probability to the lexicon of the user.
16. The non-transitory computer-readable medium of claim 15 , wherein the first value of the intra-class probability is determined based on when the utterance was received from the user.
17. The non-transitory computer-readable medium of claim 15 , wherein the first term is a name of a city and the first value of the intra-class probability is determined based on a population of the city.
18. The non-transitory computer-readable medium of claim 15 , wherein the first value of the intra-class probability is determined based on a current location of the user.
19. The non-transitory computer-readable medium of claim 18 , wherein the first term is a place and the first value of the intra-class probability is determined based on a distance between the place and the current location of the user.
20. The non-transitory computer-readable medium of claim 15 , the steps further comprising:
translating the first term from the first language to a second language,
wherein the translated first term is also added to the lexicon of the user.