Association of Terms and Phrases in Medical Field by Relative Correlation and Self Evolving Data Structure

Abstract

Since the emergence of the Internet medical knowledge is spreading around the globe increasingly fast. Though publicly available, it is a difficult task to determine individual relevance for most non-professionals. Additionally, relationships between medical terms are hard to discover even for professionals. The proposal is to build a "Clinical decision support system", which will help people to find remedies and treatments based on severity of emergencies in the form of natural language.

My goal in this project is to create self-evolving data structure based on Evolutionary algorithm. This data structure which will store medical data in a way that interlinks data so that terms and phrases will be evolutionary aligned by relative correlation by putting its medical relevancy in centre .So that we can create our own raw storage meaningful schema on which users redefined query in the form of machine reasonable and retrieve possible remedies and possible other meaning of users query.

The focus is on creating various graph based structures representing the relatedness between medical symptoms and diseases and studying various graph algorithms and applying them on structures to obtain desired suggestion a user is demanding through search engine. So Itwill be very much easy to evolve graph based structure as the volume and complexity in connectedness between medical data increases.