作者: V. Zue , S. Seneff , J. Polifroni , H. Meng , J. Glass
DOI: 10.1109/ICSLP.1996.607243
关键词: Foreign language 、 Language acquisition 、 Information access 、 Computer science 、 Human–computer interaction 、 Data structure 、 Artificial intelligence 、 Natural language processing 、 Natural language 、 Component (UML) 、 Premise 、 Context (language use)
摘要: The paper describes the authors' recent work in developing multilingual conversational systems that support human-computer interactions. approach is based on premise a common semantic representation can be extracted from input for all languages, at least within context of restricted domains. In design such systems, language dependent information separated system kernel as much possible, and encoded external data structures. internal manager, discourse dialogue component, database are maintained transparent form. They describe two possible application areas capabilities: on-line access using spoken dialogue, learning maintenance foreign system.