Automatic Addition of Verbal Semantic Attributes to a Japanese-to-English Valency Transfer Dictionary

The effectiveness of using the semantic attributes of verbs has been shown in various kinds of natural processing systems, such as machine translation systems. The addition of an attribute value is, however, time-consuming and must be performed by hand by an expert on the attribute value. In this paper, two methods for efficiently adding verbal semantic attributes to a Japanese-to-English valency transfer dictionary in a machine translation system are proposed and evaluated. One method involves a professional analyst mentally writing down decision-tree-like rules from process images when adding an attribute value to each dictionary entry. The other method involves automatically extracting a decision tree for adding attribute values from dictionary entries with semantic attribute values within a transfer dictionary using the decision tree learning program C5.0. We examine the key factors contributing towards the identification of an attribute value in the entries of the transfer dictionary. The proposed method is also applicable for adding semantic attributes effectively for dictionary entries of a bilingual dictionary within a machine translation system.