Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2001-08-28
2003-09-09
Robinson, Greta (Department: 2177)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C704S231000
Reexamination Certificate
active
06618726
ABSTRACT:
FIELD OF INVENTION
This invention relates to the general field of Internet web browsing or searching for particular web pages or other information references. More particularly, the present invention is related to speech recognition, and to identification and isolation of key words from that speech, and passing those words to web search engines or other such search functions as found on web-browsers
BACKGROUND OF INVENTION
Herein web-browsing is defined as finding information documents or web-pages on the Internet associated with a given technical or other such criteria of interest to the user. The primary mechanism to search for specific web-pages is to key in search strings of characters to a search engine or the equivalent in a commercially available browser. The searching provides a list of hits or matches, and the specific text or web-pages can be displayed. Any of the listed web pages can be brought up on the screen by known methods, e.g. “pointing and clicking” on words that are “linked” (hypertext links) to classes of information desired and bringing up those web pages on the user's screen if desired or at least bring up the text on the user's screen if graphics are not available to the user. Such web-browsing can be done in general areas or in specific topical areas. The topical areas, herein defined as micro-domains, are usually areas, like the weather, stocks, sports, finance, news, etc., in which specific search protocols with particular search requirements and formats have been developed.
Herein search facilities is defined broadly including any searching mechanism useful on the Internet. Included are specific search engines, such as AltaVista, Infoseek, Lycos, and search capabilities buried within other databases on other Internet sites.
There has been some disclosure of systems accepting voiced inputs from which searches are made. One such system, called “Galaxy”, was developed as a “natural language” system at MIT. This system is an independent browser that is not useful with the existing browsers on the market. This is a limitation since a significant amount of work must be accomplished to allow “Galaxy” to perform searches outside some specific domains for which “Galaxy” has been specifically programmed to understand.
It is an object of the present invention to provide a voice to text search string suitable for inputting into any of the commercially available search engines or databases available on the Internet.
It is yet another object of the present invention to provide web searches over specific domain names and for general searches.
Another object of the present invention is to identify key words and to identify extraneous words where the key words delivered to the search engine and the extraneous words deleted.
It is another object of the present invention to arrange the key words into search strings formatted for specific domain names.
It is yet another object of the present invention to allow search results to be re-run to identify links.
SUMMARY OF THE INVENTION
The foregoing objects are met in a method and apparatus including, means for providing search words to a search facility operating on the Internet from input voiced utterances. The voiced utterances are fed to a statistical speech recognizer where the words that were most likely voiced are determined. A database is formed of a list of words defining specific subject topics or micro-domain (see definition below) names, e.g. weather, stocks, news, and the like. The uttered words are compared to the database words. If a match is found a set of key words are specified and formatted by a subprogram associated with the topic or micro-domain. The set of key words is input to a search engine or other such search facility. The search facility or search engine returns the information about that micro-domain.
If a match to a specific topic or micro-domain is not found. a general search of the recognized words is performed. The recognized words are input to a identification module program where particular word groups are recognized. In a preferred embodiment these groups include, but are not limited to: names of people, places, organizations and companies; times; dates; percentages; and amounts of money. The key words are identified and other words are dropped. The string of search words is delivered to a search engine or other such search facility and the search engine returns the results to be displayed, preferably, on the user's monitor. The search engine may retrieve web pages found which are displayed for the user. There may be words returned that are highlighted and further searched via the identification module to be used as links for greater particularity and more detailed information.
In operation, a user speaks into electronic equipment or other such apparatus that converts the utterances into electrical signals. The utterances are passed to a speech recognizer that accepts natural continuous speech patterns and generates search words that are passed eventually to search engines. Such equipment and speech recognition software are known and in common use in the field.
Other objects, features and advantages will be apparent from the following detailed description of preferred embodiments thereof taken in conjunction with the accompanying drawings in which:
REFERENCES:
patent: 4829576 (1989-05-01), Porter
patent: 5732393 (1998-03-01), Aoshima
patent: 5802526 (1998-09-01), Fawcett et al.
patent: 5819220 (1998-10-01), Sarukkai et al.
patent: 5845074 (1998-12-01), Kobata
patent: 5915001 (1999-06-01), Uppaluru
patent: 5918223 (1999-06-01), Blum et al.
patent: 5960399 (1999-09-01), Barclay et al.
patent: 5963965 (1999-10-01), Vogel
patent: 6101473 (2000-08-01), Scott et al.
patent: 6199067 (2001-03-01), Geller
patent: 6243680 (2001-06-01), Gupta et al.
patent: 6311182 (2001-10-01), Colbath et al.
Rudnicky et al., “Speechwear: A Mobile Speech System,” pp. 538-541, vol. 1 IEEE electronic library, Oct. 3, 1996.
Madeline Bates. Models of Natural Language Understanding, Voice Communication Between Humans and Machines, David B. Roe and Jay G. Wilpon (eds.), National Academy Press, Washington, DC, pp. 238-253, 1994.
G. David Forney, Jr., The Viterbi Algorithm, Proc. of the IEEE, vol. 61, No. 3, pp. 263-278, Mar. 1973.
John Makhoul and Richard Schwartz, State of the Art in Continuous Speech Recognition, Proc. Natl. Acad. Sci. USA, vol. 92, pp. 9956-9963, Oct. 1995.
J. Robin Rohlicek et al., Continuous Hidden Markov Modeling for Speaker-Independent Word Spotting, IEEE Intl Conference on Acoustics, Speech and Signal Processing, Glasgow, Scotland, pp. 627-630, May 1989.
SRA International, Inc., Name Tag™, Product Brochure.
Charles T. Hemphill et al., Speech-Aware Multimedia, IEEE Multimedia, pp. 74-77, 1996.
Colbath Sean C.
Nguyen Long
Pham Cecile
Stallard David
Genuity Inc.
Robinson Greta
Suchyta Leonard Charles
Weixel James K.
LandOfFree
Voice activated web browser does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Voice activated web browser, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Voice activated web browser will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3025540