The publish/subscribe interaction scheme is claimed to provide the loosely coupled form of interaction required in large scale environ-ments. There are important implications for the underlying techno-logical platform. Events must be detected and signaled as notifi-cations, and they must be understandable beyond the closed confines of a single enterprise. Currently, pub/sub does not profit from the infrastructure to support such a set- ting, where data producers and consumers do not share common assumptions about exchanged data. Since pub/sub mechanisms decouple producers and consumers, they should share a common understanding in order to express their mutual interests. Events must be understandable beyond the closed confines of a single component or application (including apps that interact across traditional borders regardless of economic, cultural or linguistic differences). In this work the pub/sub paradigm is analyzed with the purpose to extend it to open distributed heterogeneous environments. On the ba-sis of this analysis, a novel approach of combining ontologies and notifications is presented to support the correct interpretation of data coming from heterogeneous sources.