Construction of Japanese Search Engine Based on Computational Model of Inductive Reasoning

Abstract

Associative search systems can be applied to the education system for language learning. In this study, we developed an information search system for language learning based on the results of previous studies. This search system is based on a computational model of inductive reasoning which uses relationships of features and nouns in structures of language knowledge (probabilistically-based language knowledge structure). The system can eliminate unwanted results in accordance with association. In addition, via the system, previous studies results were modified based on psychological experiments. Finally, a survey was conducted in order to assess the validity of the results and confirm whether the output is valid.