QUESTION ANSWERING AS DIALOGUE WITH DATA

We describe an interactive approach to question answering where the user and the system first negotiate the scope and shape of information being sought and then cooperate in locating and assembling the answer. The system, which we call HITIQA(11), has access to a large repository of unprocessed and unformatted data, and is additionally equipped with limited capabilities to search and navigate this set. The user asks questions in English, for example "What are the effects of pollution on commercial fishing on the Black Sea?" Mediating between the data and the user is a Dialogue Management System that attempts to make some sense of what the user is asking and what can be retrieved from the database, and then reconcile these through a dialogue. The purpose of the Dialogue Manager is to make the communication between the human and the machine possible and as efficient as it can be, but it does not necessarily imply a full understanding thus leaving the initiative fully in user's hands. The HITIQA system has been designed primarily for information analysts who require answers to complex, analytical questions, not just finding simple facts.