Research

Research within the ILPS group is organized around three main themes: information retrieval, language technology and semistructured data.

Much of our research is task-driven. Tasks we address concern (multi-lingual) document retrieval, question answering, novelty checking, search and analysis of user generated content, and search for experts and expertise.

In addition, we address a broad variety of extraction and shallow parsing techniques that either help address these tasks or build on the tasks listed above.

To be able to effectively and efficiently manage and retrieve the extracted information (which is semistructured in nature), a lot of research at ILPS concerns XML retrieval, web retrieval, source retrieval, ontology management (for medical and digital library applications), and query evaluation for semistructured data.

Evaluation Exercises

Much of our research is guided by, and highly dependent on, world-wide evaluation efforts of information retrieval and language processing tasks, both in terms of end-to-end tasks, and in terms of intrinsic assessments of information processing components. In particular, members of the ILPS group take part in TREC, CLEF, INEX, ACE/TERN, and Senseval. To support its evaluation efforts, the ILPS group has made, and continues to make, a considerably investment in software infrastructure.

Projects

Our research is mostly organized in externally funded projects. We have separate pages for projects.