Search-based BI is a new trend for business users to access to data analysis tools and facilitating their decision making. Since users, either are not always fully aware of the extent to which BI systems can meet their requests, or cannot explicitly express their search intention, there is a need for guiding them toward a comprehensive result. This paper presents the first steps of our research project, which aims at understanding decision makers query intention when they interact with a computer aided-decision system. The expectations of this work is to increase the levels of confidence and trust of the users in interactive BI systems, as a consequence of the feeling that they are well understood and appropriately guided.