ABSTRACT
Providing information out of data has always been the major concern in information science. Addressing and disseminating the acquired information, however, has been the major focus of research activities in query languages and question answering systems. Despite the variety of approaches taken so far, two central questions can be raised when we ask for information: how to ask for information and what kind of queries the system can answer. Although the second question has been exhaustively examined by theories and/or practical solutions referring to model-or proof-theoretic semantics, closed versus open world assumptions, fuzziness or uncertainty of query results, etc., the way how to ask for available information still remains a major challenge, especially when end-users are involved in the querying process. The roots of the challenge can be found at the assumption we make that an end-user: 1. learns the query language syntax 2. understands the meaning of the underlying database schema
in terms of adequate interpretations of data constructs such as relations, classes and collections,
3. understands the meaning of attributes and/or values, 4. is aware of the context which relates to the data as expressed in
terms of measurement units, explanation or defi nitions, etc. In this sense, the key answer to the question of how to ask
for the available information goes through the role of semantics the lack of which becomes more and more apparent in existing information systems, i.e., a semantic gap. Furthermore, the end-
user does not really know what he is looking for or how to ask for available information, unless she/he fully understands it, i.e. seeing, interpreting it. Even if the user knows what he wants, he is confronted with complex syntax formalisms, since typical systems rely only on low-level features as expressed by the data model on which the data repository relies.
