An examination of natural language as a query formation tool for retrieving information on e‐health from Pub Med

The Internet has become the first tool of choice for performing information searches. The concepts of e-health and related topics have become areas of interest for professional researchers and lay users alike, and increasingly, users turn to the Internet for information about health and health care. The purpose of this study is to examine complexity and variability of natural language as observed in using search terms that express the concept of e-health. We survey effects of this complexity in terms of precision of information retrieved, the shift in the terminology for an emerging concept and with regards to the degree of overlap between query results. We have designed a series of natural language queries and used them to perform a series of searches on Pub Med. Subsequent analysis examined the degree to which relevant information can be retrieved using terms regarding e-health. The degree to which relevant returns overlap between searches has also been examined. Our results show significant variation in relevance and overlap between different query terms and a high degree of precision among search results that have high incidence of overlap.