ToMaR -- A Data Generator for Large Volumes of Content

We present To MaR, a scalable application that supports the efficient integration of legacy applications within a MapReduce environment. The work is motivated by scenarios for scalable content processing developed in the context of the EC project SCAPE. ToMaR specifically addresses the need for extracting data sets from large volumes of binary content based on existing, content-specific applications within a scalable data management environment. This paper discusses the main functionalities of ToMaR and describes how ToMaR is utilized as part of a typical workflow. We present a real-word scenario that makes use of ToMaR for the characterization of archived web content. A workflow and experimental results which have been produced using sample content from the Web Archive Austria are discussed.