Sök:

Query expansion med hjälp av en elektronisk tesaurus i en bibliografisk online-databas.


In this Masters thesis we investigate query expansion using an electronic thesaurus. We formed 30 topics from the library and information science field. For each topic three searches were made in the bibliographic database LISA Library and Information Science Abstracts via CSA Cambridge Scientific Abstracts. LISA has an electronic thesaurus attached. We used this thesaurus to form baseline queries, baseline queries expanded with synonyms, and baseline queries expanded with both synonyms and related terms. The terms used in the baseline queries were the terms preferred by the thesaurus the "use for" terms. The three query types were analyzed with regard to precision and relative recall. The relative recall was set to 100% for the strategy where a baseline query was expanded with both synonyms and related terms. The average precision for the searches using baseline queries was 41% with a 60% relative recall. For the searches using baseline queries expanded with synonyms, the precision was 39% and the relative recall was 72%. The searches using baseline queries expanded with both synonyms and related terms resulted in an average precision of 37%. We also measured the "net effect", that is, the precision and recall with regard to the new documents retrieved by each search strategy. We found that, on average, 60% of the relevant documents were retrieved by the baseline queries. The strategy using a baseline query expanded with synonyms was the least effective one in retrieving new relevant documents on average it brought only 12% of the relevant documents. When measuring the precision for the new documents retrieved by each search strategy we encountered a methodological problem which is discussed in this thesis.

Författare

Emma Johansson Birgitta Jonsson

Lärosäte och institution

Högskolan i Borås/Institutionen Biblioteks- och informationsvetenskap (BHS)

Nivå:

Detta är en D-uppsats.

Läs mer..