Systems and methods for collaborative note-taking
First Claim
1. A method for collaborative note taking based on a speech of a speaker and providing a summary to a user in an audience of the speaker, the method comprising:
- receiving a first set of information from the speech;
performing speech recognition on the first set of information and determining selected portions of the speech;
determining portions of context information corresponding to a domain information from a presentation information source temporally associated with the selected portions of the speech;
determining at least one language model based on the selected portions of the speech and the temporally associated portions of context information from the presentation information source, wherein the at least one language model is dynamically determined;
applying the language model to the first set of information to extract salient tokens from the first set of information;
verifying relevance of the salient tokens based on the presentation information source to obtain verified tokens;
generating the summary including the extracted salient tokens, wherein generating the summary includes assembling the verified tokens;
displaying the summary to the user; and
receiving collaborative user feedback information relating to the summary and adjusting the language model according to the collaborative user feedback,wherein the method is implemented by a computer.
3 Assignments
0 Petitions
Accused Products
Abstract
Techniques are provided for determining collaborative notes and automatically recognizing speech, handwriting and other type of information. Domain and optional actor/speaker information associated with the support information is determined. An initial automatic speech recognition model is determined based on the domain and/or actor information. The domain and/or actor/speaker language model is used to recognize text in the speech information associated with the support information. Presentation support information such as slides, speaker notes and the like are determined. The semantic overlap between the support information and the salient non-function words in the recognized text and collaborative user feedback information are used to determine relevancy scores for the recognized text. Grammaticality, well formedness, self referential integrity and other features are used to determine correctness scores. Suggested collaborative notes are displayed in the user interface based on the salient non-function words. User actions in the user interface determine feedback signals. Recognition models such as automatic speech recognition, handwriting recognition are determined based on the feedback signals and the correctness and relevance scores.
244 Citations
46 Claims
-
1. A method for collaborative note taking based on a speech of a speaker and providing a summary to a user in an audience of the speaker, the method comprising:
-
receiving a first set of information from the speech; performing speech recognition on the first set of information and determining selected portions of the speech; determining portions of context information corresponding to a domain information from a presentation information source temporally associated with the selected portions of the speech; determining at least one language model based on the selected portions of the speech and the temporally associated portions of context information from the presentation information source, wherein the at least one language model is dynamically determined; applying the language model to the first set of information to extract salient tokens from the first set of information; verifying relevance of the salient tokens based on the presentation information source to obtain verified tokens; generating the summary including the extracted salient tokens, wherein generating the summary includes assembling the verified tokens; displaying the summary to the user; and receiving collaborative user feedback information relating to the summary and adjusting the language model according to the collaborative user feedback, wherein the method is implemented by a computer. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 43, 44)
-
-
20. A system for collaborative note taking based on a speech by a speaker and providing a summary to a user in an audience of the speaker, the system comprising:
-
a memory; an input/output circuit for; receiving a set of information from the speech; retrieving portions of information from the set of information; and retrieving portions of context information from a presentation information source to obtain domain information, the portions of information from the speech being temporally associated with the portions of context information from the presentation information source; a processor that performs the operations; determines at least one language model based on the portions of information from the speech and the temporally associated portion of context information from the presentation information source, wherein the at least one language model is dynamically determined; applies the language model to the set of information to extract salient tokens from the set of information; generates the summary; transmits the summary to be displayed to the user; and adjusts the language model according to the user feedback after receiving collaborative user feedback information relating to the summary; and a relevance and correctness determination circuit for verifying relevance of the salient tokens based on support information to obtain verified tokens, wherein the summary is generated by assembling the verified tokens. - View Dependent Claims (21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 45, 46)
-
-
40. Computer readable storage medium comprising computer readable program code embodied on the computer readable storage medium, the computer readable program code usable to program a computer to recognize ambiguous information comprising:
-
determining portions of information from a speech of a speaker; determining portions of context information from a presentation information source temporally associated with the portions of information from the speech; determining at least one language model based on the portions of information from the speech and the temporally associated portions of context information from the presentation information source, wherein the at least one language model is dynamically determined; applying the at least one language model to the speech to extract salient tokens from the speech; verifying relevance of the salient tokens based on the presentation information source to obtain verified tokens; determining a summary of the speech based on at least one of the determined language models, wherein determining the summary includes assembling the verified tokens; displaying the summary to a user in an audience; and receiving collaborative user feedback information relating to the summary and adjusting the language model according to the collaborative user feedback. - View Dependent Claims (41)
-
-
42. A system for recognizing information from a speaker and providing a summary for a user in an audience of the speaker, comprising:
-
means for determining portions of information from a speech of he speaker means for determining portions of context information from a presentation information source temporally associated with the portions of information from the speech, and determining domain information of the speech from the context information; means for determining at least one language model based on the portions of information from the speech and the temporally associated portions of context information from the presentation information source, wherein the at least one language model is dynamically determined; means for determining output information based on at least one of the determined language models, said means for determining output applying the language model to the speech to extract salient tokens from the speech, verifying relevance of the salient tokens based on the presentation information source to obtain verified tokens, and generating a summary of the speech by assembling the verified tokens; means for displaying the summary to the user; and means for receiving collaborative user feedback information relating to the summary and adjusting the language model according to the collaborative user feedback.
-
Specification