An extraction of medical knowledge on text mining for ubiquitous medicine

Just now, ubiquitous computing that means anyone, anyplace, and anytime we are able to use computer and those technologies exist unconsciously, is beginning. The important point concerning ubiquitous computing is interface between human and information. So, we would like to focus attention on natural language that is kind of human interface in medicine because natural language is an essential tool for healthcare professionals and patients to communicate. If natural language processing is realized, it is easy that various information is recorded anytime and anywhere and anyone with mobile, in addition, we can use data of text, speech and so on, and we are able to acquire medical knowledge from it. We have tried to extract medical knowledge with methods of text mining. To put it concretely, we have developed a new tool and medical dictionary called co-occurrence relation illustration system with medical terminology including basic property to every word. This system is able to show one of the expression and meaning of medical language. Consequently, natural language data as convenience media for input information but inconvenient media for output information to computing are useful actually and we can use computer system for medicine as talking with it ubiquitously.