Describes a method of selecting index terms directly from a word frequency list, an idea originally suggested by Goffman. Results of the analysis of word frequencies of two articles seem to indicate that the automated selection of index terms from a frequency list holds some promise for automatic indexing. (Author/MBR)

Substantial alterations to a system often have little or no effect on particular collections. This may be due to poor separation of relevant and non-relevant documents. The paper presents a procedure for characterizing this separation, which can show whether proposed modifications of the base system are likely to be useful. (8 references)…

Retrieval performance with automatic term classifications for three test collections has been variable. This paper attempts to discover why. The real difference between the collections is in the separation of relevant from non-relevant documents. The separation is so poor that classification cannot be expected to succeed. (14 references)…

Describes large scale experiments to test the value of index term weighting based on relevance information, and presents results showing that weighting leads to remarkable improvements in retrieval performance. (Author/CWM)

The Latent Semantic Indexing and MatchPlus systems examine similar contexts in which words appear and create representational models that capture the similarity of meaning of terms and then use the representation for retrieval. Text Retrieval Conference experiments using these systems demonstrate the computational feasibility of using…

Presents an algorithm that automatically extracts words and phrases containing geographic place names or characteristics from a text document and uses them as input to database functions that use spatial reasoning to approximate statistically the geoposition referenced in the text. Preliminary results and future work are discussed. (28 references)…