Intention-Based Information Retrieval of Electronic Health Records

Abstract

The large volume of information stored in electronical health records is very valuable in the medical field, e.g., for clinical research and administrative purposes. However, health care professionals still face difficulties to recover and select relevant data. Although literature has investigated the influence of lexical, syntactical and semantic parameters in information retrieval techniques, few studies explore intentions as explicit users' actions in information recovery. This article studies means of considering intentions in search engines. We define a method with an original algorithm to properly rank search results from a set of electronical health records. This investigation relies on well-established theories to categorize several types of intentions. The proposed technique is based on knowledge representation models for annotating meanings and intentions in medical records. Achieved results are illustrated in scenarios based on real medical cases.