Development of search engine for Medline database with ranking of search results from perspective of evidence-based medicine and Grade criteria

Mikhail Kamalov

Abstract: Nowadays there is a number of search engines which conform of a wide range of user queries to the largest corpus of bibliographic medical records MEDLINE. However, all existing search engines do not solve the problem of retrieving and ranking medical articles from the evidence-based medicine perspective. The work describes a process of search engine implementation with ranking of search inference from the levels of evidence. Experiments with classification of medical studies by clinical/nonclinical, intervention/observation, by subtype of medical interventions, and by levels of evidence are conducted. Novel algorithm of ranking medical studies based on levels of evidence is described.