Variations on KWIC

KWIC—Key Word In Context—is a form of automatic indexing using computers. It is automatic in the sense that the computer system determines the indexing from free text input. Luhn first described the method in which text of a length which can be accommodated on a single line of print is indexed at the centre of the page, the words that precede and follow the indexing keyword being displayed on the same line. As a control device a list of words for which no indexing entry should be made is stored in the computer program. This list, often referred to as a ‘stop list’, includes the articles, conjunctions and commonly occurring trivial words which do not form useful indexing entries. Less frequently occurring ‘useless’ entries are carried, but at little cost and with little damage to the usefulness of the index. The system is often referred to as quick and dirty: quick in that it is cheap to run, easy to prepare and not demanding of intellectual decisions at input; dirty in that useless entries will appear in the index, and, since the indexing vocabulary is uncontrolled, the user must consider alternative words that may express the concept for which he is searching. When the input is prepared in‐house a number of control features can be used to supplement the stop list to make the index more effective but requiring more thought at input. This paper concerns a number of such added control features.