Contents
Next:
List of Figures
Up:
Identifying and Merging Related
Previous:
Identifying and Merging Related
Contents
Acknowledgments
List of Figures
List of Tables
Introduction
What is a digital library?
Integrating bibliographic databases
Overview of related work
Networked information discovery and retrieval
Libraries and cataloging
Database systems
Information retrieval
Cataloging and the Computer Science Collection
Bibliographic relationships and cataloging
Issues in cataloging and library science
Taxonomy of relationships between records
Expressing relations between Bibtex records
Practical issues for identifying relationships
Bibtex record format
CS-TR record format
MARC records
Effects of cataloging practices
Describing the contents of DIFWICS
Subjects covered in collection
Other characteristics of collection
Identifying Related Records
Record comparison in the presence of errors
Algorithm for author-title clusters
Comparing author lists
String comparisons with n-grams
Performance of algorithm
Other systems for identifying related records
Duplicate detection in library catalogs
The database merge/purge problem
Analysis of author-title clusters
Merging Related Records
Goals of merger and outline of process
Creating the union record
Problems with union records
Refinements to merger process
Clusters sizes and composition in DIFWICS
Presenting Relations and Clusters
The basic Web interface
Assessing the quality of union records
Automatic linking
Searching for related Web citations
Principles for fully-automated system
Conclusions and Future Directions
Future directions
Performance, portability, and production
Improving the quality of records
Identifying other bibliographic relationships
Integrating non-bibliographic information
Enabling librarianship and human input
Conclusions
References
About this document ...
Jeremy A Hylton
Mon Feb 19 15:33:12 1996