A methodology for authoring information aggregation and retrieval workflows in scalable infrastructures

With the increasing popularity of stream-based information sources on the Web, a seemingly continuous and endless flow of messages is generated and propagated to the users. This makes it important to provide systems for reducing the information load on users by employing filtering and ranking techniques. Constructing such systems is an important task in the research domain of Internet Information Retrieval (IIR). In this paper we are presenting a novel approach for creating processing workflows for information aggregation and retrieval tasks. It is backed by a system architecture called Spectre, which allows the realization of distributed and highly scalable systems. A graphical authoring tool named Spectre Cockpit allows defining, configuring and deploying IIR systems providing an all-embracing methodology for the IIR authoring process.