Empirical significance of document partitions is investigated as a function of index term-weight and similarity thresholds. Results show the same empirically preferred partitions can be detected by two independent strategies: an analysis of cluster-based retrieval analysis and an analysis of regularities in the underlying structure of the document…

Describes an expected precision (EP) measure designed to predict document retrieval performance. Highlights include decision theoretic models; precision and recall as measures of system performance; EP graphs; relevance feedback; and computing the retrieval status value of a document for two models, the Binary Independent Model and the Two Poisson…

The Hepatitis Knowledge Base (text of prototype information system) was used for modifying and testing "A Navigator of Natural Language Organized (Textual) Data" (ANNOD), a retrieval system which combines probabilistic, linguistic, and empirical means to rank individual paragraphs of full text for similarity to natural language queries…

Considers classifications produced by application of single linkage, complete linkage, group average, and word clustering methods to Keen and Cranfield document test collections, and studies structure of hierarchies produced, extent to which methods distort input similarity matrices during classification generation, and retrieval effectiveness…

This update of 1978 bibliometric study of online retrieval literature reveals that: literature has continued to grow, reaching peak annual output (504 papers) in 1981; conference literature has become centralized in four major proceedings; online retrieval papers have appeared in 479 journals; and 31 authors have contributed 10 or more papers.…

Defines relationship between user profile and user query in terms of relationship between clusters of documents retrieved by each, and explores the expression of cluster similarity and cluster overlap as linear functions of similarity existing between original pairs of profiles and queries, given the desired retrieval threshold. (23 references)…

Reports algorithm for calculation of term discrimination values that is sufficiently fast in operation to permit use of exact values. Evidence is presented to show that relationship between term discrimination and term frequency is crucially dependent upon type of inter-document similarity measure used for calculation of discrimination values. (13…

Zipfian laws for frequency distributions of word pairs and longer phrases are derived from text sample analysis. From crossing of Zipfian curves, it is deduced that number of multi-word phrases that occur frequently in text is surprisingly small, of same order of magnitude as number of individual word-types. (8 references) (EJS)

Reports on study that used set of measures for characterizing performance with different databases and set of hypotheses for predicting search difficulties based on cognitive processes involved in decision-making aspects of information search. Highlights include confidence assessment, system design criteria, transparency, metatransparency, and…