Characteristics of Fulltext Index by Human and Automatic Indexing Systems

The purpose of this study is to investigate the characteristics of indexes by human and machine, and differences between them in terms of term identification in a full text environment. A back-of-book index and two indexes produced by two term identifiers (LinkIt and Termer) as pseudo-indexing systems for a whole body of a monograph are examined. In the investigation, the traditional contrast between manual and automatic indexing is confirmed in full text environment: manual index is for browsing and human use, and automatic index is for searching and machine use. The border between them. however, becomes vague. Some considerations for the use of the term identifiers for browsing and for searching are discussed, and further research for the use of the term identifier is suggested.