Automated indexing

Automated indexing

An automatic computer indexing program can only produce a list of words and phrases occurring in the text (a concordance).

It cannot

1. Make an intelligent analysis of a text - so that readers interested in (for example) population are also led to discussions on demography.

2. Recognize relationships with other terms, by providing appropriate cross references
            eg. literacy see also reading.

3. Distinguish between homographs (words that look the same but have different meanings)
           eg. records (best performances)
                 records (audio recordings)

4. Provide access to terms for users who search under synonyms by providing cross references
           eg. pupils see students

5. Split long strings of page numbers into useful groups under subheadings, or consolidate a number of similar word variations (challenge/challenges/challenged/challenging).

Generally an automatically produced index will not be very user-friendly. Computer full-text searching also has many of these drawbacks, and cannot replace the usefulness of an index.