In biomedical domains, high throughput technologies pro- duce large amount of transcriptomic data used for studying comportment of genes. The analysis and the interpretation of such data require impor- tant databases and e cient mining methods, in order to extract speci c biological functions belonging to a group of genes of an expression pro le. To this aim, we propose here a new approach for mining transcriptomic data combining domain knowledge and classi cation methods. Firstly, we propose the de nition of Fuzzy Di erential Gene Expression Pro les (FD-GEP) based on fuzzy classi cation and a di erential de nition be- tween the considered biological situations. Secondly, we will use our pre- viously de ned e cient semantic similarity measure (called IntelliGO), that is applied on Gene Ontology (GO) annotation terms, for comput- ing semantic and functional similarities between genes of the resulting FD-GEP and well known genetic markers involved in the development of cancers. After that, the similarity matrices will be used to introduce a novel Functional Spectral Representation (FSR) calculated through a semantic ranking of genes regarding their similarities with the tumoral markers. The FSR representation should help expert to interpret by a new way transcriptomic data and infer new genes having similar biolog- ical functions regarding well known diseases. Availability: The semantic similarity measure