Toward a Model of Interaction for Complex Search Tasks

Information on the Web and in digital libraries has rich annotation but the current Search engines excel at short queries that fail to exploit this power. The potential of semantic annotation is not realized in shallow navigational search but holds the promise to significantly enhance complex tasks or information needs, by supporting the task as a whole over multiple interactions with the system. We analyse the different phases of interaction for a complex search task, and give insight in different goals of the interaction in each the phase. The resulting model of interaction suggests ways in which the semantic annotation can be exploited to iteratively articulate the information need and to explore the search results.