Phonetic searching

• US 7,769,587 B2
• Filed: 11/26/2008
• Issued: 08/03/2010
• Est. Priority Date: 03/04/2000
• Status: Expired due to Term

• Current Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Original Assignee: Georgia Tech Research Corporation (University System of Georgia)
• Inventors: Price, William E., Cardillo, Peter S., Clements, Mark A.
• Primary Examiner(s): Sked; Matthew J
• Application Number:
  US12/323,601
• Publication Number:
  US 20090083033A1
• Time in Patent Office: 615 Days
• Field of Search: None
• US Class Current: 704/243
• CPC Class Codes: G06F 16/3343 using phonetics
• CPC Class Codes: G06F 16/685 using automatically derived transcript of audio data, e.g. lyrics speech recognition G10L15/00
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

An improved method and apparatus is disclosed which uses probabilistic techniques to map an input search string with a prestored audio file, and recognize certain portions of a search string phonetically. An improved interface is disclosed which permits users to input search strings, linguistics, phonetics, or a combination of both, and also allows logic functions to be specified by indicating how far separated specific phonemes are in time.

Claims

1. A method comprising:

processing, using a computer, a text source to form a multi-dimensional array of elements, wherein each element is indexed by at least a first dimension representative of a time frame of an audio representation of the text source and a second dimension representative of a phoneme state, and further wherein each element comprises a value that characterizes a probability associated with an occurrence of a phoneme state within a time frame of the audio representation of the text source;

receiving a search query via a text-based interface of a computer, an audio-based interface of a computer, or some combination thereof;

processing the query using a computer to generate a phonetic representation of the query; and

searching, using a computer, the multi-dimensional array of elements according to the phonetic representation of the query to locate one or more putative instances of the query in the text source.

2. The method of claim 1, wherein processing the query comprises determining whether the query is linguistically specified and if so, converting the linguistically specified query into a sequence of phonemes.

3. The method of claim 1, wherein processing the query comprises determining whether the query is phonetically specified and if so, generating a sequence of phonemes from the phonetically specified query.

4. The method of claim 1, further comprising:

generating a display that indicates, in sorted order, where in the text source the query is most likely to occur.

5. The method of claim 1, wherein the search query comprises a voice command received via the audio-based interface.