Necessary tools choice in a particular situation for computer conversation

Abstract

Human beings make associations based on their general knowledge, with intellectual thought activity forming the basis of natural conversation. When human beings converse, they first absorb what the other party is saying, then use their analogical and associative abilities to continue the conversation based on naturally acquired general knowledge. Such general knowledge-based conversation is difficult to simulate using the language data and methods that have been acquired thus far. Therefore, it is necessary to provide machines with the intellectual structure they require to understand the semantic connections between words, determine how these words relate to one another, and render sensible judgments. In this paper, we propose a method for constructing a system aimed at determining what tools are necessary in a particular situation based on the intellectual structure of sensible judgment. For example, this system would examine data input like "cut the cabbage" as it appears in a conversational sentence, and associate it with relevant tools such as "kitchen knife" and "cutting board". In this manner, rather than by simply parsing the phrase as linguistic data, the system allows for responsive dialogue that is relevant to the situation. © 2011 Springer-Verlag.