Analysis, generation and semantic representation in CONTRAST—a context‐based machine translation system
Hitoshi Isahara, Uchida Yuriko · Systems and Computers in Japan · 1995
Abstract This paper describes both a Japanese language analysis system and an English language generation system. The knowledge utilized by these systems has been incorporated into a context‐based machine translation system (CONTRAST) developed by the Electrotechnical Laboratory (ETL). The Japanese language analysis component of the CONTRAST system consists of individual modules to analyze syntax, semantics and context. The syntactic analysis module is a completely parallel parsing system based on Earley's algorithm. The semantic analysis module conducts analysis based on case relations. The contextual analysis module creates an intermediate representation that correspond to original text. It does this by placing events included in the semantic analysis results for each individual sentence within a framework that is read out from the general knowledge described as a process model in the concept dictionary. The English text generation system of CONTRAST begins with this intermediate representation. Then, referring to the concept dictionary (common with the analysis module), it creates English text according to generation rules created from the real life text data dependent knowledge relating to English newspaper articles. The format of the concept dictionary, which consists of descriptions of general knowledge, and the format of the intermediate representation, which is a description of the contextual information contained in each body of text, are the same. The outline of the conceptual system employed and studies on the process of creating this system are also discussed.