The research in this dissertation consists of two parts: An active learning algorithm for hierarchical labels and an embedding-based retrieval algorithm. In the first part, we present a new approach for learning hierarchically decomposable concepts. The approach learns a high-level classifier (e.g., location vs. non-location) by separately…

The diversity of engineering education research provides an opportunity for cross-fertilisation of ideas and creativity, but it also can result in fragmentation of the field and duplication of effort. One solution is to establish a standardised taxonomy of engineering education terms to map the field and communicate and connect research…

Surveillance video recording is becoming ubiquitous in daily life for public areas such as supermarkets, banks, and airports. The rate at which surveillance video is being generated has accelerated demand for machine understanding to enable better content-based search capabilities. Analyzing human activity is one of the key tasks to understand and…

The concept of tri-citation is introduced, as a logical extension of co-citation, and a geometrical model (the circle) is devised to account for these and all other forms of multiple citation. (Author)

Discusses the need to retrieve ideas in addition to content retrieval by such means as classification, subject headings, and keywords. Three models for idea retrieval are described, and a new model is proposed which combines selection, a controlled vocabulary, and logical structure. (seven references) (CLB)

Outlines a cognitive process model of abstracting, indexing, and classification that is based on text comprehension processes. Text comprehension for indexing versus other purposes is discussed, including conceptual and perceptual processing; conceptual knowledge and the development of expertise are discussed; and characteristics of short-term and…

Explores the use of linear models and a combination of neural networks and linear classifiers to create a hybrid hierarchical mixture of experts (HME) model. Results confirm that using the hierarchical structure of the classification vocabulary improves categorization performance. (AEF)

Proposes an automated document classification system based on the Gray code, information-theoretic considerations, and requirements of classification systems including objectivity, inclusion, linearity, and increasing distance-dissimilarity. A measure to evaluate system performance is developed, and the proposed system is compared with the Library…

Presents an information systems navigation approach using a combination of functionalities encountered in classification processes, Hypertext systems, and information retrieval systems. The first version of the SYRIUS system is described, with emphasis on the user's graphic interface which integrates these functionalities. (26 references) (EA)

Examines the literature on the myth of "hard science-soft science" as a means of defining scientific specialties and classifying science literature, and discusses some shortcomings of scholarship in library and information science. Use of Toulmin's categorization of sciences to organize the literature and provide access for scientists is…

This paper begins with some general remarks about the indexing of electronic publications. Characteristics of today's information world are highlighted, including decentralization, distributed data collections, and heterogeneous data structures and indexing procedures. Three possible models for managing access to information are summarized: a…

Proposes a model that assists in understanding indexing on the World Wide Web and presents an experiment assessing the validity of Inverse Document Frequency (IDF) as a term weighting strategy for Web documents. Findings indicate that IDF scores are not stable in the heterogeneous and dynamic context of the Web. (Author/AEF)

Describes the development of a software system, SIMPR (Structured Information Management: Processing and Retrieval), that will process documents by indexing them and classifying their subjects. Topics discussed include information storage and retrieval, file inversion techniques, modelling the user, natural language searching, automatic indexing,…

Describes the development and outcome of a cooperative computerization effort involving nine public and university museums. The discussion covers the project goals, data development and analysis, and contributions resulting form the project. Appendices provide the system data fields used and a comparison between automation of library and museum…

Explores creating a thesaurus from a cognitive viewpoint, based on a system that organizes its knowledge representation/classification according to the authors' and users' images of the subject domain, in order to increase user-text interaction. Discusses the general framework, methods for implementing author-text structures, and an approach…