Abstract: A method for creating and maintaining ontologies. The method includes transmitting a two dimensional table to a requestor. The table includes domain concepts and semantic primitives related to a domain. A data value in the table at an intersection of a domain concept and a semantic primitive indicates that the domain concept is characterized by the semantic primitive. A command to modify the table is received from the requestor. The table is updated in response to the command. The updated table is stored as an ontology.

Abstract: A method for location-based notifications, the method including receiving a request for a notification from an input device. A current location of the input device is determined in response to the receiving. A reminder location is set responsive to the current location of the input device determined in response to the receiving. On a periodic basis, a new current location of the input device is determined and compared to the reminder location. A notification is provided if the new current location is within a proximity threshold of the reminder location.

Abstract: A method for reducing speech recognition errors. The method includes receiving an N-best list associated with a user utterance. The N-best list includes one or more hypotheses and associated confidence values. The user utterance is classified in response to the N-best list, resulting in a classification. A re-scoring algorithm that is tuned for the classification is selected. The re-scoring algorithm is applied to the N-best list to create a re-scored N-best list. A hypothesis for the value of the user utterance is selected based on the re-scored N-best list.

Abstract: A method for reducing recognition errors. The method includes receiving an N-best list associated with an input of a computer based recognition system. The N-best list includes one or more hypotheses and associated confidence values. The input is classified in response to the N-best list, resulting in a classification. A re-scoring algorithm that is tuned for the classification is selected. The re-scoring algorithm is applied to the N-best list to create a re-scored N-best list. A hypothesis for the value of the input is selected based on the re-scored N-best list.