Literature/1977/Soergel


 * http://www.dsoergel.com/cv/An%20Automated%20Encyclopedia%20%20a%20Solution%20of%20the%20Information%20Problem.

Authors

 * University of Maryland

Abstract
Due to redundancy and lack of specific access, in formation transfer through the literature is costly in time to the user; it is also costly to the author who must repeat much context information before coming to his new contribution. A non-redundant data store or automated encyclopedia could alleviate these problems. The structure of such a data store and procedures for its creation from the literature by controlled removal of redundancy are described. However, an automated encyclopedia would create problems of its own. Its structure would be enormously complex and it would be very expensive to establish. Its acceptance by users and authors is by no means a given. Pilot projects in high-use areas such as medicine and drugs, or statistical methods, and in numerical data compilation are suggested to test the feasibility of an automated encyclopedia and determine a realistic scope for it.

Contents

 * 1) Introduction
 * 2) The Problem
 * 3) A solution: the automated encyclopedia
 * 4) Reasons that lead to redundancy in publications in the present system
 * 5) Use of an automated encyclopedia. Its role in information transfer
 * 6) Selective dissemination of data
 * 7) Retrospective retrieval of data
 * 8) Types of requests
 * 9) Data analysis and inference
 * 10) Forms of output
 * 11) Contributing data to the store
 * 12) Structure of an automated encyclopedia
 * 13) Logical structure of the data store
 * 14) Representation of data elements
 * 15) Data reduction through generalization
 * 16) Context indexing of data elements
 * 17) Reliability
 * 18) Indexing of potential uses/applications
 * 19) Source indications
 * 20) Remarks on physical storage organizations
 * 21) Building and updating the data store
 * 22) Introduction
 * 23) Procedures for building a non-redundant data store from the literature
 * 24) The principles of collocation of information and controlled redundancy removal
 * 25) Procedure for processing one document for inclusion of its data elements into the data store
 * 26) Problems arising in removing redundancy and establishing the logical structure of the data store
 * 27) Keeping track of the sources
 * 28) What constitutes a data element?
 * 29) Data element "modes"
 * 30) Editing sections of text expressing a data element
 * 31) Terminological standards
 * 32) Sameness of meaning and data element variants
 * 33) General discussion of the determination of same ness of meaning
 * 34) Preliminary list of possible relationships between natural language representations of data elements
 * 35) Completeness of coverage vs. summarization
 * 36) The principle of complete (non-selective) coverage and evaluative indexing
 * 37) Social and political aspects of an automated encyclopedia
 * 38) Motivation for contribution and use
 * 39) Regulation of access for contributors
 * 40) Regulation of access for users
 * 41) Other problems
 * 42) Implementation of an automated encyclopedia system
 * 43) Should an automated encyclopedia system be implemented?
 * 44) Costs
 * 45) Benefits
 * 46) Cost-benefit comparison
 * 47) Implementation of pilot projects
 * 48) Preparation for an automated encyclopedia in the present system
 * 49) Conclusions