Information Retrieval is a research area that has gained attention over thepast two decades. Few of these researches have taken place in the biomed-ical domain where satisfying users’ information needs are relatively difficultto be met. The goal of this project is to find out if it is possible to usestatistical methods in Biomedical Information Retrieval (IR) and improveretrieval performance, i.e. finding ways of fulfilling user information needs,in the biomedical domain using clustering with knowledge from the BioTracerproject.K-Mean and Expectation Maximization (EM) approaches to clustering havebeen implemented in this project with more emphasis on the EM. Both ap-proaches are used to re-ranking users searched results in an attempt to findways of fulfilling their information needs. Comparison between the Expec-tation Maximization and the K-mean are drawn in terms of their retrievalperformance i.e. precision and recall, the performance of EM compared to ex-isting approaches to search results re-ranking using clustering and problemsfaced while implementing the EM.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:ntnu-13634 |
| Date | January 2011 |
| Creators | Yeboah, Stephen |
| Publisher | Norges teknisk-naturvitenskapelige universitet, Institutt for datateknikk og informasjonsvitenskap, Institutt for datateknikk og informasjonsvitenskap |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.002 seconds