7. CONCLUSION
Three models supporting information retrieval were covered, with a particular emphasis on their mode of representation of the documents and their processing algorithms. As the oldest model the Boolean Retrieval model has the advantage of simplicity and convenience. It is however restrictive in the formulation of similarities. Similarity can be in two states only: true or false. At the other extreme, the probabilistic approach is an attempt to model the subjective features of the information retrieval process over a range of probabilities. The calculation of probabilities requires the specification of assumptions that can be highly biased and inconsistent. The document representation in the probabilistic model is very simple and ignores terms frequency or position. The Vector Space Model on the other hand combines clarity with flexibility. The underlying algebraic model is well specified and well understood, and the documents are represented with more details. The VSM offers a viable compromise in information retrieval processing.