Interactive analysis of datacube, in which a user navigates a cube with a sequence of queries to find and understand unexpected data, is often tedious. To better support this process, we propose in this paper to connect two techniques proposed earlier in this domain. These techniques are, on the one hand, discovery driven analysis, that guides the user towards regions of the cube they will find of interest, and on the other hand, query recommendation, that takes advantage of what the other users did during former analyses. Benefiting from these techniques we propose a framework for recommending OLAP queries to the user by taking into account what previous users found of interest and the explanation they worked out.