Presents a technique for dividing text into field-coherent passages based upon extracting field-associated words or phrases from the text by determining how topics grow, shrink and shift from sentence to sentence. Proposes measures of topic continuity and transition and suggests how those may be used to find the passage boundaries. (Author/LRW)

Discussion of information filtering in information retrieval systems introduces a method for defining filters for structured text where the text structure is originally defined by a grammar consisting of a set of productions. Use of a template that is created interactively from the grammar is explained, and a prototype implementation is described.…

Describes in detail the practice of information retrieval within business organizations based on 46 interviews. Highlights include the importance of text to the performance of information work rather than technical aspects of retrieval; information management; and information retrieval problems in organizations. (LRW)

Describes the design and construction of Resume Automatique a Fragments Indicateurs (RAFI), a system of automatic text summary which sums up scientific and technical texts. The RAFI system transforms a long source text into several versions of more condensed texts, using discourse analysis, to make searching easier; it could be adapted to the…

Presents a new application for retrieving news from a large electronic bank of newspapers that is intended to manage past issues of newspapers. Highlights include a data model for newspapers, including metadata and metaclasses; document definition language; document retrieval language; and memory organization and indexes. (Author/LRW)

Presents a single framework for conducting topic analysis that performs both topic identification and text segmentation. Key characteristics of the framework are: representing a topic by means of a cluster of words closely related to the topic; and employing a stochastic model, called a finite mixture model, to represent a word distribution within…

Text similarity measurements are used to determine relationships between natural-language texts and text excerpts. The resulting linked hypertext maps can be broken down into text segments and themes used to identify different text types and structures, leading to improved information access and utilization. Examples are provided for text…

Discusses the nature of term groupings, phrases, and text windows in full-text documents and computes the statistical significance of windows. Topics include classifying documents within disciplines or on a theory versus practice spectrum; and grammatical characteristics for automatic classification of documents, for information retrieval, and for…

Describes SALOMON (Summary and Analysis of Legal texts for Managing Online Needs), a system which automatically summarizes Belgian criminal cases to improve access to court decisions. Highlights include a text grammar represented as a semantic network; automatic abstracting; knowledge acquisition and representation; parsing; evaluation, including…