Examines statistical techniques for exploiting relevance information to weight search terms. These techniques are presented as a natural extension of weighting methods using information about the distribution of index terms in documents in general. (Author)

Presents notations and definitions necessary to identify the concepts and relationships that are important in modelling information retrieval objects and processes in the context of vector spaces. Earlier work on the use of vector models is evaluated in terms of the concepts introduced and certain problems are identified. (Author/EM)

Instead of leaving the user at the point where he or she receives the information from the reference storage and retrieval system, we should follow him or her to see what the impact is from the information received, how it actually affects the user's performance in his or her task. (Author/PF)

The inverse relationship of recall and precision is not a fundamental law. It must not be used as an excuse for failing to bring about any overall improvement in the performance of information retrieval systems (12 references) (Author/NH)

A hypothetical methodology is presented for measuring retrieval effectiveness by using the personal utility, based on a user's subjective evaluation, of a retrieval system's output. Because the methodology is impractical, compromise methods are outlined and their underlying assumptions, which can serve as criteria for formulating appropriate…

The difficulties of readily obtaining peripheral information and drawbacks of citation searches are discussed. A citation survey of inorganic chemistry journals is compared with previous studies. The publication of a chemical newspaper is proposed. (11 references) (Author)

The relevance assessments belonging to the Cranfield II document/query collection are shown to be faulty, in the sense that many" relevant documents were not so identified by the Cranfield judges. 9 references. (Author)

The indexing of Cranfield I and II is given and critical comments made of this indexing. Comparisons of Cranfield indexing to other types of indexing are made. (AB)

The complete report, ERIC document ED 022 494, is abstracted in the February 1969 "Research in Education.

Addresses the application of automatic classification methods to the problems associated with computerized document retrieval. Different kinds of classifications are described, and both document and term clustering methods are discussed. References and notes are provided. (Author/JD)

The present experiments were designed to study the effects of search term weighting based on very limited relevance information, supplied, for example, by one or two relevant documents. The tests simulated iterative searching as in an on-line system and show that even very little relevance information can be extremely valuable. A list of…

Discusses information retrieval techniques used on the World Wide Web. Topics include machine learning in information extraction; relevance feedback; information filtering and recommendation; text classification and text clustering; Web mining, based on data mining techniques; hyperlink structure; and Web size. (LRW)

Proposes a method to improve retrieval performance of the vector space model by using users' relevance feedback. Discusses the use of singular value decomposition and the latent semantic indexing model, and reports the results of two experiments that show the effectiveness of the proposed method. (Author/LRW)

This study examined the contribution to retrieval effectiveness (measured by recall and precision) of adding content-bearing information such as abstracts and tables of contents to bibliographic records. It was found that the addition of content-bearing information improves overall retrieval effectiveness. Improvement was primarily in terms of…

Describes a study that used the TREC information-retrieval test collection to evaluate Visual Information Retrieval Interfaces (VIRIs). Discusses multiple-similarity measures, scaling properties, and MLE (maximum likely method), and suggests that cosine-vector and overlap measures for similarity appear to recover optimal data relationships among…