RETRIEVING APPARATUS, RETRIEVING METHOD, AND COMPUTER PROGRAM PRODUCT

• US 20090083029A1
• Filed: 02/29/2008
• Published: 03/26/2009
• Est. Priority Date: 09/25/2007
• Status: Active Grant

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Koga, Toshiyuki, Suzuki, Kaoru, Yamamoto, Koichi, Doi, Miwako
• Granted Patent: US 8,374,845 B2
• Time in Patent Office: Days
• US Class Current: 704/200
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/088 Word spotting

Abstract

A word coinciding with a key word input by speech and a word related to the word are set as retrieval candidate words based on a word dictionary in which words representing formal names and aliases of the formal names are registered in association with a family attribute indicating a familiar relation among the words. Content related to any one of retrieval words selected out of the retrieval candidate words and a word related to the retrieval word is retrieved.

Claims

1. A retrieving apparatus comprising:

a first storing unit that stores a content;

a second storing unit that stores a word dictionary in which a plurality of words are registered in association with readings of the words and each of the words representing a formal name and an abbreviated name of the formal name is registered in association with a family attribute among the words;

a speech input unit that receives an input of a speech signal;

a speech recognizing unit that recognizes a character string corresponding to the speech signal as a keyword;

a retrieval-candidate expanding unit that retrieves a word coinciding with the keyword and a word familiar with the word from the word dictionary as a retrieval candidate word;

a retrieval-word selecting unit that receives a selection designation that designates any of words included in the retrieval candidate words as a retrieval word;

a retrieval-word expanding unit that sets the retrieval word and a word familiar with the retrieval word as expanded retrieval words, based on the word dictionary; and

a retrieving unit that retrieves a content related to any of words included in the expanded retrieval words from the first storing unit.

2. The apparatus according to claim 1, wherein the retrieval-word expanding unit sets only a word representing a formal name among the word coinciding with the keyword and the word familiar with the word, as the retrieval candidate word.

3. The apparatus according to claim 1, further comprising:

a registering unit that registers each of the words extracted from the content and a presence attribute indicating a storage location of the content including the each of the words in association with the word dictionary, wherein

the retrieving unit retrieves a word associated with the presence attribute among the words included in the expanded attribute words, from the storage location indicated by the presence attribute.

4. The apparatus according to claim 3, further comprising:

an external communication unit that acquires related information related to a registered word registered in the word dictionary, from information resources present on a network; and

an extracting unit that extracts a word that is in a familiar relation with the registered words from a character string included in the related information, wherein

the registering unit registers the word extracted by the extracting unit in the word dictionary in association with a family attribute indicating the familiar relation with the registered word.

5. The apparatus according to claim 4, wherein

the extracting unit extracts a reading of the word that is in the familiar relation with the registered word from the character string included in the related information, and

the registering unit registers the reading of the words extracted by the extracting unit in the word dictionary in association with a word corresponding to the reading.

6. The apparatus according to claim 4, further comprising:

a third storing unit that stores a connection destination table in which a field of an object represented by each of the words and a presence location of the information resources corresponding to the fields are associated with each other, wherein

the external communication unit acquires the related information from the presence location of the information resources corresponding to the field of the object represented by the registered word, based on the connection destination table.

7. The apparatus according to claim 1, wherein

the first storing unit stores, as the content, a content material reproducible constantly and content information describing information concerning content reproducible at predetermined date and time, and

the retrieving unit retrieves content including any of words included in the expanded retrieval words from the first storing unit.

8. A retrieving method comprising:

receiving an input of a speech signal;

recognizing a character string corresponding to the speech signal as a keyword;

retrieving a word coinciding with the keyword and a word familiar with the word from a word dictionary as a retrieval candidate word, the word dictionary registering a plurality of words in association with readings of the words and each of words representing a formal name and an abbreviated name of the formal name in association with a family attribute among the words;

selecting any of words included in the retrieval candidate words as a retrieval word;

setting the retrieval word and a word familiar with the retrieval word as expanded retrieval words, based on the word dictionary; and

retrieving a content related to any of words included in the expanded retrieval words.

9. A computer program product having a computer readable medium including programmed instructions for retrieving a content, wherein the instructions, when executed by a computer, cause the computer to perform:

receiving an input of a speech signal;

recognizing a character string corresponding to the speech signal as a keyword;

retrieving a word coinciding with the keyword and a word familiar with the word from a word dictionary as a retrieval candidate word, the word dictionary registering a plurality of words in association with readings of the words and each of words representing a formal name and an abbreviated name of the formal name in association with a family attribute among the words;

selecting any of words included in the retrieval candidate words as a retrieval word;

setting the retrieval word and a word familiar with the retrieval word as expanded retrieval words, based on the word dictionary; and

retrieving a content related to any of words included in the expanded retrieval words.