michael strube

I lead the Natural Language Processing (NLP) Group at HITS gGmbH, Heidelberg, Germany. There, I am involved in NLP related projects, work with the computational linguists at HITS, and supervise PhD. students.

(This isn't really news ...) According to Google Scholar the AAAI '06 and AAAI '07 papers Simone Paolo Ponzetto and I co-authored are the most cited papers of the top conference in AI (Google Scholar query for AAAI 2006 and AAAI 2007)!

The JAIR paper Simone Paolo Ponzetto and I published in 2007 is the runner up (i.e., we'll receive the 2010 Honorable Mention) for the the 2010 IJCAI-JAIR Best Paper Prize! For the blurb, check here.

Paper on coreference resolution at COLING '10 published (19% acceptance rate for oral presentations): Jie Cai and Michael Strube: End-to-End Coreference Resolution via Hypergraph Partitioning (PDF). We describe a system which performs coreference resolution globally in one go without separating the task into classification and clustering (decoding).

Paper on coreference resolution evaluation metrics at SIGdial '10 accepted: Jie Cai and Michael Strube: Evaluation Metrics For End-to-End Coreference Resolution Systems. We describe evaluation metrics which can deal with system mentions.

New project in the NLP group:
CoSyne: A project on Multi-lingual Content Synchronization with Wikis -- an ICT-STREP project funded by the European Commision. The project started in March 2010.

Tutorial accepted at the 11th Pacific Rim International Conference on Artificial Intelligence, Daegu, Korea, August 30 - September 3, 2010 (PRICAI '10). The topic is Extracting Knowledge from Wikipedia for Semantic Processing.

Together with Anette Frank and Stefan Riezler I lead the PhD colloquium in the CL Department at the University of Heidelberg. We also have a couple of external presentations.

I teach a class on Natural Language Generation in the CL Department at the University of Heidelberg in the summer term 2010.

Linguistics:

• Text and Dialogue
• Pragmatics

Computational Linguistics:

• Coreference Resolution
• Anaphora and Deixis in Spoken Dialogue
• Generation of Referring Expressions
• Models of Attentional State
• Discourse and Dialogue Structure (though I don't believe in it)
• Lexical Semantics
• Hedge Detection

Natural Language Processing:

• Information Extraction
• Knowledge Acquisition, Ontology Learning
• Automatic Summarization
• Detecting Vague and Manipulative Language
• Natural Language Generation Systems
• Multi-modal Dialogue Systems

Publications in Journals and Books, Conference Proceedings, Workshop Proceedings, and Complete List of Publications

Search for my publications at Google Scholar.

Send me email if you want to get a copy of a paper not linked on these pages (in those cases we had to transfer the copyright to the respective publishers; maybe linguists should follow the good example of JAIR and start to publish in open access journals).

A Few Recent Publications

• Cai, Jie; Strube, Michael (2010).
  End-to-End Coreference Resolution via Hypergraph Partitioning
  In: COLING '10, pp.143-151. (PDF)

• Nastase, Vivi; Strube, Michael (2009).
  Combining Collocations, Lexical and Encyclopedic Knowledge for Metonymy Resolution
  In: EMNLP '09, pp.1219-1224. (PDF)

• Filippova, Katja; Strube, Michael (2008).
  Sentence Fusion via Dependency Graph Compression
  In: EMNLP '08, pp.177-185. (PDF)

• Nastase, Vivi; Strube, Michael (2008).
  Decoding Wikipedia Categories for Knowledge Acquisition
  In: AAAI '08, pp.1219-1224. (PDF)

• Ponzetto, Simone Paolo; Strube, Michael (2007).
  Knowledge Derived from Wikipedia for Computing Semantic Relatedness
  In: Journal of Artificial Intelligence Research 30, pp.181-212. (PDF).
  Runner up for the IJCAI-JAIR best paper prize 2010 (awarded to an outstanding paper published in JAIR in the preceding five calendar years).

• Filippova, Katja; Strube, Michael (2007).
  The German Vorfeld and Local Coherence
  In: Journal of Logic, Language and Information, 16(4), pp.465-485.
  (locked up at Springer's page, accessible maybe to you, but not to me)

• Ponzetto, Simone Paolo; Strube, Michael (2007).
  Deriving a Large Scale Taxonomy from Wikipedia
  In: AAAI '07, pp.1440-1445. (PDF)

• Filippova, Katja; Strube, Michael (2007).
  Generating Constituent Order in German Clauses
  In: ACL '07, pp.320-327. (PDF)

• Strube, Michael; Ponzetto, Simone Paolo (2006).
  WikiRelate! Computing Semantic Relatedness Using Wikipedia.
  In: AAAI '06, pp.1419-1424. (PDF)
  This paper did not receive the best paper award (though it was nominated for it), but appears to be the most cited paper of the 2006 conference (out of 236 published papers)!

I received my Ph.D. from the (now defunct) Computational Linguistics Department at the University of Freiburg, Germany, in December 1996 under the supervision of Udo Hahn. Between 1997 and 1999 I was a postdoctoral fellow at the Institute for Research in Cognitive Science at the University of Pennsylvania, Philadelphia, PA. In 2000 I joined the European Media Lab in Heidelberg, Germany, as a researcher. Since 2001 I am group leader of the Natural Language Processing (NLP) Group of HITS, an institute which rapidly underwent several transformations before (hopefully) arriving at its final destination:

European Media Lab → EML Research → HITS.

Literature

What is Jazz?

Running

Photography