Data processing: speech signal processing – linguistics – language – Linguistics – Natural language
Reexamination Certificate
1999-10-21
2002-11-19
Edouard, Patrick N. (Department: 2654)
Data processing: speech signal processing, linguistics, language
Linguistics
Natural language
C704S255000
Reexamination Certificate
active
06484136
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is related to language recognition and more particularly to language recognition by multiple connected computer systems connected together over a network.
2. Background Description
Speech recognition systems for voice dictation, typically, try to improve word recognition accuracy by using what is referred to, normally, as a “local context” for word recognition. Local context based word recognition uses n preceding and m succeeding words, normally referred to as “n-tuples,” to identify a particular spoken word. The local context recognition model selects the most likely word based on the statistical probability of the specific n-tuples. However, these prior speech recognition systems still make inappropriate recognition errors because they lack global context for the spoken words.
For example in dictating a review of a conference paper, prior art speech recognition systems have recognized “in terms of validity” as “in terms of the lead Dave.” “Validity” has a global context because it is commonly discussed in conference paper reviews. However, “the lead Dave” is nonsensical in this context and so, has no global context.
Thus, there is a need for improved methods and systems for language recognition.
SUMMARY OF THE INVENTION
It is a purpose of the invention to improve language recognition by computers;
It is yet another purpose of the invention to expand the context base for language recognition.
The present invention is a language recognition system, method and program product for recognizing language based input from computer users connected together over a network of computers. Each computer includes at least one user based language model trained for a corresponding user. The language models may be used by users in automatic speech recognition, handwriting recognition, machine translation, gesture recognition or other similar actions that require interpretation of user activities. Computer users on a network of computers are clustered into classes of similar users. Users that are connected over the network are clustered according to their characteristic similarities such as, nationality, profession, sex, age, or any other characteristics that may influence user language model selection. Characteristics of users are collected by sensors and from databases and, then, distributed over the network during user activities. Language models with similarities among similar users on the network are identified.
The language models include a language model domain, with similar language models being clustered according to their domains. Language models identified as similar are modified in response to user production activities. After modification of one language model, other identified similar language models are compared and adapted. Also, user data, including information about user activities and user language model data, is transmitted over the network to other similar users. Language models are adapted only in response to similar user activities, when these activities are recorded and transmitted over the network. Language models are given a global context based on similar users that are connected together over the network.
REFERENCES:
patent: 5426583 (1995-06-01), Mendibil
patent: 5444617 (1995-08-01), Merialdo
patent: 5510981 (1996-04-01), Berger et al.
patent: 5659765 (1997-08-01), Nii
patent: 5664058 (1997-09-01), Vysotsky
patent: 5768603 (1998-06-01), Brown et al.
patent: 5897616 (1999-04-01), Kanevsky et al.
patent: 5907821 (1999-05-01), Kaji et al.
patent: 5950157 (1999-09-01), Heck et al.
patent: 6088669 (2000-07-01), Maes
patent: 6161083 (2000-12-01), Franz et al.
patent: 6163769 (2000-12-01), Acero et al.
patent: 6182038 (2001-01-01), Balakrishnan et al.
patent: 6253179 (2001-06-01), Beigi et al.
Kanevsky Dimitri
Wolf Catherine G.
Zadrozny Wlodek W.
Edouard Patrick N.
Percello Louis J.
Peterson, Jr. Charles W.
LandOfFree
Language model adaptation via network of similar users does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Language model adaptation via network of similar users, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Language model adaptation via network of similar users will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2928496