Data processing: speech signal processing – linguistics – language – Speech signal processing – Recognition
Reexamination Certificate
1999-10-05
2002-08-13
Korzuch, William (Department: 2741)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Recognition
C704S275000, C704S010000, C704S009000, C707S793000
Reexamination Certificate
active
06434524
ABSTRACT:
BACKGROUND
Description of the Related Art
As computers have become more prevalent it has become clear that many people have great difficulty understanding and communicating with computers. A user must often learn archaic commands and non-intuitive procedures in order to operate the computer. For example, most personal computers use windows-based operating systems that are largely menu-driven. This requires that the user learn what menu commands or sequence of commands produce the desired results.
Furthermore, traditional interaction with a computer is often slowed by manual input devices such as keyboards or mice. Many computer users are not fast typists. As a result, much time is spent communicating commands and words to the computer through these manual input devices. It is becoming clear that an easier, faster and more intuitive method of communicating with computers and networked objects, such as web-sites, is needed.
One proposed method of computer interactions is speech recognition. Speech recognition involves software and hardware that act together to audibly detect human speech and translate the detected speech into a string of words. As is known in the art, speech recognition words by breaking down sounds the hardware detects into smaller non-divisible sounds called phonemes. Phonemes are distinct units of sound. For example, the word “those” is made up of three phonemes, the first is the “th” sound, the second is the “o” sound, and the third is the “s” sound. The speech recognition software attempts to match the detected phonemes with known words from a stored dictionary. An example of a speech recognition system is given in U.S. Pat. No. 4,783,803, entitled “SPEECH RECOGNITION APPARATUS AND METHOD”, issued Nov. 8, 1998, assigned to Dragon Systems, Incorporated. Presently, there are many commercially available speech recognition software packages available from such companies as Dragon Systems, Inc. and International Business Machine Corporation.
One limitation of these speech recognition software packages or systems is that they typically only perform command and control or dictation functions. Thus, the user is still required to learn a vocabulary of commands in order to operate the computer.
A proposed enhancement to these speech recognition systems is to process the detected words using a natural language processing system. Natural language processing generally involves determining a conceptual “meaning” (e.g., what meaning the speaker intended to convey) of the detected words by analyzing their grammatical relationship and relative context. For example, U.S. Pat. No. 4,887,212, entitled “PARSER FOR NATURAL LANGUAGE TEXT”, issued Dec. 12, 1989, assigned to International Business Machines Corporation teaches a method of parsing an input stream of words by using word isolation, morphological analysis, dictionary look-up and grammar analysis.
Natural language processing used in concert with speech recognition provides a powerful tool for operating a computer using spoken words rather than manual input such as a keyboard or mouse. However, one drawback of a conventional natural language processing system is that it may fail to determine the correct “meaning” of the words detected by the speech recognition system. In such a case, the user is typically required to recompose or restate the phrase, with the hope that the natural language processing system will determine the correct “meaning” on subsequent attempts. Clearly, this may lead to substantial delays as the user is required to restate the entire sentence or command. Another drawback of conventional systems is that the processing time required for the speech recognition can be prohibitively long. This is primarily due to the finite speed of the processing resources as compared with the large amount of information to be processed. For example, in many conventional speech recognition programs, the time required to recognize the utterance is long due to the size of the dictionary file being searched.
An additional drawback of conventional speech recognition and natural language processing systems is that they are not interactive, and thus are unable to cope with new situations. When a computer system encounters unknown or new networked objects, new relationships between the computer and the objects are formed. Conventional speech recognition and natural language processing systems are unable to cope with the situations that result from the new relationships posed by previously unknown networked objects. As a result, a conversational-style interaction with the computer is not possible. The user is required to communicate complete concepts to the computer. The user is not able to speak in sentence fragments because the meaning of these sentence fragments (which is dependent on the meaning of previous utterances) will be lost.
Another drawback of conventional speech recognition and natural language processing systems is that once a user successfully “trains” a computer system to recognize the user's speech and voice commands, the user cannot easily move to another computer without having to undergo the process of training the new computer. As a result, changing a user's computer workstations or location results in wasted time by users that need to re-train the new computer to the user's speech habits and voice commands.
SUMMARY
The embodiments of the present invention include a novel and improved system and method for interacting with a computer using utterances, speech processing and natural language processing. Generally, the system comprises a speech processor for searching a first grammar file for a matching phrase for the utterance, and for searching a second grammar file for the matching phrase if the matching phase is not found in the first grammar file. The system also includes a natural language processor for searching a database for a match entry for the matching phrase; and an application interface for performing an action associated with the matching entry if the matching entry is found in the database.
In one embodiment, the natural language processor updates at least one of the database, the first grammar file and the second grammar file with the matching phrase if the matching entry is not found in the database.
The first grammar file is a context-specific grammar file. A context-specific grammar file is one that contains words and phrases that are highly relevant to a specific subject. The second grammar file is a general grammar file. A general grammar file is one that contains words and phrases which do not need to be interpreted in light of a context. That is to say, the words and phrases in the general grammar file do not belong to any parent context. By searching the context-specific grammar file before searching the general grammar file, the present invention allows the user to communicate with the computer using a more conversational style, wherein the words spoken, if found in the context specific grammar file, are interpreted in light of the subject matter most recently discussed.
In a further aspect of the present invention, the speech processor searches a dictation grammar for the matching phrase if the matching phrase is not found in the general grammar file. The dictation grammar is a large vocabulary of general words and phrases. By searching the context-specific and general grammars first, it is expected that the speech recognition time will be greatly reduced due to the context-specific and general grammars being physically smaller files than the dictation grammar.
In another aspect of the present invention, the speech processor searches a context-specific dictation model for the matching phrase if the matching phrase is not found within the dictation grammar. A context-specific dictation model is a model that indicates the relationship between words in a vocabulary. The speech processor uses this to determine help decode the meaning of related words in an utterance.
In another aspect of the present invention, the natural language processor replaces at least one word in t
Azad Abul K.
Korzuch William
One Voice Technologies, Inc.
Pillsbury & Winthrop LLP
Sze James Y. C.
LandOfFree
Object interactive user interface using speech recognition... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Object interactive user interface using speech recognition..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Object interactive user interface using speech recognition... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2947592