Our research aim is to ameliorate the recall of the Fuzzy Proximity Information Retrieval Model ( FPIRM ) of Beigbeder & Mercier ( 2005) , their approac h is very "precise" when evaluating internationally agreed upon collections of documents used for benchmarking. The precision in this case for each query is the ratio of relevant documents retrieved over documents returned. However, the recall is weak. By recall we mean the ratio of relevant documents retrieved over all relevant documents in the collection (Rijsbergen, 1979) . FPIRM approach evaluates the relevance between a document and a query by using a fuzzy fun ction that takes into account the distance between the occurrences of query terms in a document. Our research studies the use of semantic query expansion to increase the recall of their model.