Method for Interacting With Users of Speech Recognition Systems

• US 20090216538A1
• Filed: 02/25/2008
• Published: 08/27/2009
• Est. Priority Date: 02/25/2008
• Status: Active Grant

• Current Assignee: Mitsubishi Electric Research Laboratories, Inc. (Mitsubishi Electric Corporation)
• Original Assignee: Mitsubishi Electric Research Laboratories, Inc. (Mitsubishi Electric Corporation)
• Inventors: Harsham, Bret A., Schmidt-Nielsen, Bent, Ramakrishnan, Bhiksha, Weinberg, Garrett
• Granted Patent: US 7,917,368 B2
• Time in Patent Office: Days
• US Class Current: 704/275
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning

Abstract

A computer implemented method facilitates a user interaction via a speech-based user interface. The method acquires spoken input from a user in a form of a phrase of one or more words. It further determines, using a plurality of different domains) whether the phrase is a query or a command. If the phrase is the query the method retrieves and presents relevant items from a plurality of databases. If the phrase is a command, the method performs an operation.

Claims

1. A computer implemented method for interacting with a user via a speech-based user interface, comprising the steps of:

acquiring spoken input from a user in a form of a phrase of one or more words;

determining, using a plurality of different domains, whether the phrase is a query or a command; and

retrieving and presenting relevant items from a plurality of databases if the phrase is the query; and

otherwise

performing an operation if the phrase is a command.

2. The method of claim 1, in which the phrase includes variables and modifiers if the phrase is the command.

3. The method of claim 1, in which the phrase is in a form of particles, and an index to the items is in the form of the particles.

4. The method of claim 1, in which the relevant items are segregated on a domain-by-domain basis.

5. The method of claim 1, in which the relevant items are presented in a relevance order.

6. The method of claim 1, in which the relevant items commingled in a combined result list.

7. The method of claim 1, in which all possible items form a set of items, and further comprising:

grouping items in the set of into subsets of items, in which each subset of items forms one of the plurality of domains;

retrieving the items from one or more of the plurality of domains in response to the query; and

selecting a particular domain as most relevant to the query if the retrieved items are least likely to have been obtained by random draw from the entire set of items.