Compares two similarity measures for information retrieval in a document vector space, the distance-based and the angle-based, and presents a newly developed similarity measure based on both the distance and angle strengths of two compared objects. Introduces the concept of the iso-extent contour and then compares all three similarity measures.…

Proposes a new approach for classifying text documents into two disjoint classes. Highlights include a brief overview of document clustering; a data mining approach called the One Clause at a Time (OCAT) algorithm which is based on mathematical logic; vector space model (VSM); and comparing the OCAT to the VSM. (Author/LRW)

Presents a new corpus-based method for constructing a thesaurus based on lexical cooccurrence. Each word is represented as a vector in a multi-dimensional space that captures cooccurrence, and a similarity measure compares these vectors. Two methods for using these thesaurus vectors in information retrieval are described. (Author/LRW)

Presents the human evaluation of ILIAD, a program for machine-aided indexing that was designed to assist expert librarians in computer-aided indexing and document analysis. Topics include controlled indexing and free indexing; natural language and concept-based information retrieval; evaluation methodology; syntactic variations; and a comparison…

Examines two models to study the consequences for coding efficiency of sample fluctuations, a specific type of error in the statistics on which data compression codes are based. The possibility that random fluctuation can be exploited for data compression is discussed. (six references) (KRN)

Considers relevance and recall as measures of the effectiveness of information retrieval systems. Highlights include relevance judgments and information retrieval, including experiments with test collections; information retrieval and assessment of recall; and a comparison of two retrieval mechanisms using a new set of relevance judgments designed…

Considers the consistency in selection of search concepts and search terms carried out by human intermediaries in online bibliographic retrieval. Degree of consistency is calculated on the basis of search terms and concepts both for one individual over time and by comparing the performance of several individuals on one request. (JKP)

Critically contrasts the three main schools of thought on knowledge and assesses the resulting implications for knowledge management. Presents a conceptual model to integrate theoretical and practical themes to serve as a framework for developing a future research agenda for the development of knowledge management business tools and applications.…

Discusses hierarchic document clustering in information retrieval and its potential improved effectiveness over inverted file search (IFS). Describes experiments conducted to determine whether hierarchic clustering can increase the retrieval effectiveness compared to both static clustering and of conventional IFS when it is applied to search…

Discusses reuse of existing software for new purposes as a key aspect of efficient software engineering by matching formal written requirements used to define the new and the old software. Explores two matching methodologies that use information retrieval techniques and describes test results from a comparison of two military systems. (Author/LRW)

Describes a method for compressing tomographic images obtained using Positron Emission Tomography (PET) and Magnetic Resonance (MR) by applying transform compression using all available dimensions. This takes maximum advantage of redundancy of the data, allowing significant increases in compression efficiency and performance. (13 references) (KRN)

Examines previous research on keyword-search techniques and evaluates a keyword classification scheme. Compiled a list of 1056 management information systems (MIS) keywords from 14,676 articles in 27 (MIS) journals and compared it to an existing list (Barki). Concludes that a comprehensive keyword-classification system is needed for enhanced MIS…

Investigates the durability of accounting research in scholarly journals by evaluating the extent and usage of previous literature in current literature and ranks 49 journals in accounting research by citation count and time analysis. Found the mode, median, and mean number of years of citations to be three, seven, and nine years. (PEN)

Proposes a new method of text classification using stochastic decision lists, ordered sequences of IF-THEN-ELSE rules. The method can be viewed as a rule-based method for text classification having advantages of readability and refinability of acquired knowledge. Advantages of rule-based methods over non-rule-based ones are empirically verified.…

Discussion of word frequency and keywords in text focuses on a method to estimate automatically the stability classes that indicate a word's popularity with time-series variations based on the frequency change in past electronic text data. Compares the evaluation of decision tree stability class results with manual classification results.…