Most classification tasks that have been tackled for automation are ones involving complex chains of causal reasoning. However, our daily lives are filled with simple classification tasks. In order to provide an easy-to-use and unified interface for various simple classification tasks, a natural language front-end is introduced. This natural language interface supports a lexicon-based solution to the problem of automatic classification. When a user provides a natural language input, whether it be a full sentence or a simple memo consisting of a sequence of noun phrases, the natural language interface generates a parsed structure. By analyzing the parser output, a head word is extracted from an input expression. A semantic hierarchy associated with the head, drawn from a lexical database, makes the automatic classification possible via inheritance. Although a classical organization of a lexical database constitutes the backbone for our system, we found that its information content was far from being sufficient. A few augmentations to the existing lexical database has also been suggested. Our experiment shows the approach resulted in an automatic classifier with about 92% accuracy.
Natural Language Understanding, Intelligent User Interface, Automatic Classification, Thesaurus, Semantic Inheritance.