Fuzzy Systems

In the previous chapter, we saw an overview of the theory and techniques for building intelligent systems that are capable of processing natural language input. It is certain that there will be a growing demand for machines that can interact with human beings via natural language. In order for the systems to interpret the natural language input and react in the most reasonable and reliable way, the systems need a great degree of fuzziness. The biological brain can very easily deal with approximations in the input compared to the traditional logic we have built with computers. As an example, when we see a person, we can infer the quotient of oldness without explicitly knowing the age of the person. For example, if we see a a two-year-old baby, on the oldness quotient, we interpret the baby as not old and hence young. We can easily deal with the ambiguity in the input...