Presents an efficient text categorization (or text classification) algorithm for document retrieval of natural language texts that generates bigrams (two-word phrases) and uses the information gain metric, combined with various frequency thresholds. Experimental results suggest that the bigrams can substantially raise the quality of feature sets.…

Discussion of natural language processing focuses on term weighting in information retrieval. Presents a new weighting method that depends on low frequency terms, called negative weighted inverse verb frequency, and discusses case frames, word similarity, similarity measurement, and recall and precision improvement. (LRW)

Discusses general problems of thesaurus construction theory and practice. Highlights include lexical control and its tools in various databases; natural language versus conceptual networks; systems approach; thesaurus versus classification, including associative relations; and thesaurus role in information storage and retrieval. (110 references)…

Describes research conducted at the TREC (Text Retrieval Conference) interactive track that compared Boolean and natural language searching, showing they achieved comparable results; and assessed the validity of batch-oriented retrieval evaluations, showing that the results from batch evaluations were not comparable to those obtained in…

Discussion of natural language processing focuses on a multi-strategic integrated text preprocessing method for difficult problems of sentence boundary disambiguation and word boundary disambiguation of Web texts. Describes an evaluation of the method using Korean Web document collections. (Author/LRW)

Presents an approach to summarization that combines information from multiple facts into a single sentence using linguistic constructions. Describes two applications: one produces summaries of basketball games, and the other contains summaries of telephone network planning activity. Both summarize input data as opposed to full text. Discusses…

Presents a new extraction pattern, modified Document Type Definition (mDTD), which relies on analytical interpretation to identify extraction target from the contents of Web documents. Experiments with 330 Korean and 220 English Web documents on audio and video shopping sites yielded an average extraction precision of 91.3% for Korean and 81.9%…

Discusses two query evaluation strategies used in large text retrieval systems: (1) term-at-a-time; and (2) document-at-a-time. Describes optimization techniques that can reduce query evaluation costs. Presents simulation results that compare the performance of these optimization techniques when applied to natural language query evaluation. (JMV)

Describes an information retrieval system in which advanced natural language processing is used to enhance the effectiveness of term-based document retrieval by preprocessing the documents; discovering interterm dependencies and build a conceptual hierarchy specific to database domain; and processing the user's natural language requests into…

Describes and evaluates a system that selects key information from an event database by reasoning about event frequencies, frequencies of relations between events, and domain-specific importance measures. The system aggregates similar information and plans a summary tailored to a stereotypical user. (AEF)

Increasing information retrieval performance using phrases and part-of-speech (POS) information is one example of a type of decision-making performance that is improved when using this linguistic information. The relative effectiveness of using multi-term phrases as opposed to individual terms is shown, as well as the relative worth of POS tagged…

Explores the indirect method of locating for indexing the likely explicit and implicit captions of photographs, using multimodal clues including the specific words used, syntax, surrounding layout of the Web page, and general appearance of the associated image. The MARIE-3 system thus avoids full image processing and full natural-language…