Printed indexes were the precursors to hypertext links. If you have good indexing skills, you can apply those skills to writing indexes for either printed books or online documents. Although locator systems are different in electronic media than in printed books, the basic principles of indexing apply to both online documents and back-of-the-book (b-o-b) indexes. Most online indexes look very much like b-o-b indexes; however, because online information is not linear, the biggest difference is that hypertext links in online documents serve the same purpose as See and See also cross-references in b-o-b indexes. Another difference is that most indexes for online documents use just one or, at the most, two levels of index entries--that is, main headings and subentries, but no sub-subentries.
Indexing online information is a key skill for online information developers, yet not much information exists on how to do it right. After defining online indexing and briefly presenting the research the authors were involved in, the article provides all the key ingredients for creating a good, searchbased online index.
To deal with AOL's size, I contracted others to help me with 'the dirty work' of typing search words in the database records. At the height of the project, four people worked in a large room at open desks. Although I was in charge of the project, most of the nitty gritty was accomplished by two other individuals. Following my lead, they reviewed each of the AOL pages, decided the important concepts of each area, and chose representative vocabulary. With their assistance and the involvement of several Songline employ- ees, the project took under one year, from the initial planning stages through testing, review, and summary. In theory, then, AOL's size could be conquered by allowing enough time to complete the project and contracting enough indexers to do the work.
Online indexing has great potential as a tool for information retrieval, although current online indexes are not always well used. Research and experience indicate that online indexing can be most effective if it is approached as a combination of traditional indexing and using computer search capabilities. Typical search facilities have great power but tend to rely on complex algorithms or else retrieve more information than users can effectively sort through. Traditional indexing techniques serve as a filter for concepts to limit searches to information that users will actually find valuable. To take full advantage of search facilities, online indexes can be designed with a flat (nonhierarchical) structure in which each index entry is clearly worded and makes use of keywords from the subject matter. Indexers can include additional keywords as synonyms that point to the relevant index entries. When indexers take advantage of these concepts and when index users clearly understand what to expect from online indexes, the the indexes become an extremely powerful retrieval medium.
Online indexes have the same logical structure as print indexes with main headings and usually subheadings. Some online indexes can also be searched electronically. A search request in Yahoo! returns a list of online category headings. Online indexes, like their print cousins, are true 'searchable structures,' not simply concordance lists of terms appearing in the text.