Enhanced services for targeted information retrieval by event extraction and data mining

Abstract

We present a framework combining information retrieval with machine learning and (pre-)processing for named entity recognition in order to extract events from a large document collection. The extracted events become input to a data mining component which delivers the final output to specific user's questions. Our case study is the public collection of minutes of plenary sessions of the German parliament and of petitions to the German parliament. © 2008 Springer-Verlag Berlin Heidelberg.