Fuzzy DCG Syntactic Parser for Command Language Recognition under Special Conditions

J.L. Ortiz and M. Valle (Puerto Rico)


Artificial Intelligence Applications, Natural LanguageProcessing, Fuzzy Logic, Intelligent Parser


The application of speech technology to allow people to interact with computers in advanced simulation systems or real life applications by using automatic speech recognition (ASR) is becoming very important. ASR is the capability of a computer to convert spoken language to recognized words in textual form. Speech signals for commands and control messages must be processed in a reliable fashion especially under adverse conditions including noisy environments, different speaker accents, and stress. The conversion from spoken language to text can generate mistakes due to several environmental and human conditions that “confuse” conversion algorithms resulting in a wrong text output. The fuzzy syntactic parser presented in this paper helps detect and correct these errors in the text output. Humans are able to understand commands under special or adverse conditions due to their experience, common sense, and other cognitive abilities. This paper presents a fuzzy syntactic parser implemented using a definite clause grammar model and Lukasiewicz logic system to overcome these difficulties. A prototype model in Prolog is presented.

Important Links:

Go Back