The relevance of a document with respect to a query depends on many factors that are very difficult to model in an exact way. Not surprisingly, the probabilistic approach is so far the most successful approach. It is based on the assumption that the distribution of the indexing features will tell us something about the relevance of a document. In this section we introduce a suitable probability space and the corresponding terminology to model IR events (Fuhr 1992b).
Unable to display preview. Download preview PDF.