LarKC is an EU FP7 project and aims at developing a platform for reasoning on massive heterogeneous information such as social media data. The role of Politecnico and Cefriel, forming a JRU (Joint research Unit) within LarKC, is addressing Stream Computing, i.e. supporting querying and reasoning tasks on top of RDF Streams. Stream-based data sources such as sensors, feeds, click streams, and stock quotations have become increasingly important in many application domains. Streaming data are received continuously and in real-time, either implicitly ordered by arrival time, or explicitly associated with timestamps. As it is typically impossible to store a stream in its entirety, Data Stream Management Systems (DSMS) allow one to register continuously running queries that return new results as new data flow within the streams. At the same time, reasoning upon very large RDF data collections is advancing fast, and SPARQL has gained the role of standard query language for RDF data. Also, SPARQL engines are now capable of querying integrated repositories and collecting data from multiple sources. Still, the large knowledge bases now accessible via SPARQL are static, and knowledge evolution is not adequately supported. There is a strong need for the convergence of SPARQL engines and stream data management.