Application of type-2 fuzzy logic to healthcare literature search at point of care

The biomedical field publishes a huge volume of articles every year and most of them are now available online as PDF full-texts. here is still no an effective search mechanisim that could locate the full-text articles very narrowly matching the users preference quickly. Such a mechanisim can be important for many real-time applications such as information at point of care for a physician who only has a minute or two to find relevant information in the literature for his/her decision making. We explore to develop such a mechanism with the help of type-2 fuzzy sets and fuzzy logic. Our current focus is on the full-text articles in the PubMed database. We have implemented the system using Java programming language. We show the preliminary result on the usefulness of the fuzzy logic. Unlike most search engines, which are interactive in nature, our eventual system is designed to operate automatically (hence, it can be used for literature monitoring, for instance).