Describes the development of a method for the automatic identification of sublanguage vocabulary words as they occur in abstracts. Highlights include research relating to sublanguages and their vocabulary; domain terms; evaluation criteria, including recall and precision; and implications for natural language processing and information retrieval.…

The traditional concern with relevance or nonrelevance of unexamined documents is misplaced, and traditional measures of effectiveness should be replaced by estimates of direct utility of examined documents. (Author)

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…

Describes techniques used for information retrieval from free-text or document databases and presents a new approach based on signature files, or representations of documents, called the Map Array Signature (MAS) method. False drops are discussed, and results of experiments are compared to Robert's Superimposed Coding method. (20 references) (LRW)

Investigates the presence of clustering structure in a document collection and the influence of the presence of clustering structure on the success of cluster-based retrieval. Term-weight and similarity thresholds are discussed, empirical and statistical significance are considered, and indexing exhaustivity for document representation is…

Describes a conceptual model for bibliographic data called Maestro (Management Environment for Structured Text Retrieval and Organization) that is being developed both to support structured documents in databases, and to provide a query language to link information contained in these structures. Traditional text retrieval systems are discussed,…

Describes the development of a structural model of document retrieval based on lexical-semantic relationships between index terms. An algorithm that uses tree-to-tree distance to measure structural closeness between a document and a query statement is explained, and the proposed model is compared to a vector retrieval model. (18 references) (LRW)

Presents the "Evidence-of-Existence" information system in which the structure is developed, with application of cybernetic concepts, as an isomorphic model in analogy to the system structure of communication technology. Three criteria of structuring are postulated: (1) source-channel-sink, with input-output characteristics, (2) filter-type…

This issue contains nine articles that provide an overview of trends and research in parallel information retrieval. Topics discussed include network design for text searching; the Connection Machine System; PThomas, an adaptive information retrieval system on the Connection Machine; algorithms for document clustering; and system architecture for…

Describes significant attributes of expert systems, contrasts them to conventional computer systems, and provides an overview of the R1 expert system used by Digital Equipment Corporation (DEC) to put together operational systems that meet customers' requirements. Document handling, particularly pictures and images in documents, is also briefly…

Reviews aspects of the relationship between machine learning and information retrieval. Highlights include learning programs that extend from knowledge-sparse learning to knowledge-rich learning; the role of the thesaurus; knowledge bases; artificial intelligence; weighting documents; work frequency; and merging classification structures. (78…

Summarizes various linguistic approaches proposed for document analysis in information retrieval environments. Topics discussed include syntactic analysis; use of machine-readable dictionary information; knowledge base construction; the PLNLP English Grammar (PEG) system; phrase normalization; and statistical and syntactic phrase evaluation used…

These two articles introduce a new model of probablistic information retrieval (PIR) that is based on the statistical technique of the combination of expert opinion (CEO), which in this model is the opinions of indexers and users. Part 1 presents the conceptual basis, while Part 2 explains the mathematical details. (73 references) (LRW)

Describes a model for the workload of a full text retrieval system and of a simulation model for a version of an existing system called Full/Text. The effectiveness of various data caching strategies using the 2 models are explored, and the impact of various parameters on the strategies are evaluated. (13 references) (LRW)

This discussion of the structure of online databases and the structure of related subject disciplines focuses on terms used in a book about tropical diseases and terms used in the MEDLINE database for articles on these diseases. Matrices used are explained and Medical Subject Heading (MeSh) terms are examined. (16 references) (LRW)