



In particular, in our implementation the i-th item is represented by a vector Vi(g1. For this reason, a visualization tool proposed by the authors to support user navigation in conceptual spaces has been modified to support user navigation in neurobiological hyperspaces populated by the original items to better understand the associations discovered by the mining tool. Therefore, interacting visualization environments to allow the user to access the relevant original items are envisaged. Namely, the resulting associations are expressed by formulas such as Gene-Disease pairs or Gene-Disease-Protein triplets, whereas the user does not receive any support to retrieve the items which may offer an interpretative key to understand the novel relationships pointed out by the mining tool. Generally, all the available mining tools are not provided with interactive visualization shells. Not all the identified associations were explicitly contained in some item, but many of them derived from inference chains. In previous papers, the authors developed a tool for mining medical documents for extracting relevant associations.
