A question answering system on evidence based medicine

A QUESTION ANSWERING SYSTEM ON EVIDENCE BASED MEDICINE Alirıza Arıbaş, Medical Informatics MSc Program, Inciraltı Campus Izmir, ali.aribas@ceng.deu.edu.tr ABSTRACT Objective: In this study, it is aimed to design and implement a question answering system that can search semantics in PubMed data, according to PICO format, which will enable researchers in health to keep themselves updated despite of increasing number of scientific articles. Method: PubMed database was downloaded and then imported to the database. Semantic triple relations - subject, predicate and object - are extracted from the abstracts and saved into the database. User questions were taken as PICO formatted questions and transformed into SQL query to fetch results from the database. The result was shown as a list descending order by citation count and completion date of the article. Results: Useful information was retrieved by PICO formatted question from the extracted information. Conclusion: A question answering system, which is developed for finding appropriate evidence in PICO format from the PubMed archive, is achieved and available to use publicly. Keywords: PubMed, question answering, evidence based medicine, PICO