A steadily increasing number of natural language (NL) documents are handled in information systems. Most of these documents typically contain some formatted data, which we call strong database data, and additionally some unformatted data, i.e., free text. The task of a modern information system is to characterize such unformatted (text) data automatically and, in doing so, to support the user in storing and retrieving natural language documents. The retrieval of natural language documents is a fuzzy process because the user will formulate fuzzy queries unless he uses some strong search keys. Retrieval of natural language documents can be facilitated with natural language queries; that is, with searches based on natural language text comparisons.
Teufel, Bernd, "NATURAL LANGUAGE DOCUMENTS: INDEXING AND RETRIEVAL IN AN INFORMATION SYSTEM" (1988). ICIS 1988 Proceedings. 24.