An Information Retrieval System for Improving Efficiency in Scientific Literature Searches

Conducting scientific research most often involves a search through existing literature in order to avoid repeating research efforts, review methods already developed for solving a problem, gain a better understanding of a problem, etc. Typically, this search is performed using the Internet, which is a convenient portal to various databases of books, journal articles, technical reports, preprints, etc. In using this approach for information retrieval (IR), a researcher has the advantage of being able to search through great amounts of reference material. However, along with this great access comes the challenge of efficiently retrieving and processing only relevant material. When using an information retrieval engine to search through electronic resources, simple queries can return too many documents or documents not relevant to the intended search criteria. For example, if a physicist is researching which methods have been used to solve problems in the area of plasma physics, a search for “methods plasma physics” on the World Wide Web using Google (www.google.com) yields more than 27,000 documents (out of approximately 2.5 billion). The physicist could never hope to read through all of those documents. Even searching arXiv (xxx.lanl.gov), a server for preprint articles in physics, for this same information on plasma physics yields more than 350 documents (out of approximately 135,000). Still, this is too much information to process. Just reading the abstracts of those 350 articles would require a great deal of time, and there is no guarantee that this document set would be representative enough to constitute a thorough literature search. With these issues in mind, the goal of this project is to develop an information retrieval system that