Welcome to the Wiki of the seminar ''Information Extraction'' in the winter semester 2013 / 2014

This seminar will be given by Elmar Haußmann and Sabine Storandt, from the chair of Prof. Dr. Hannah Bast (who will also attend the sessions). The seminar will take place every Wednesday, 4:00 pm - 6:00 pm ct., in the seminar room 00-007 in building 106. There will be no session on Wednesday, December 25, 2013, on Wednesday, January 1, 2014 (christmas break) and on Wednesday, October 30, 2013.

Information Extraction is the task of extracting (structured) information from natural language text. We will cover interesting topics in this area with a focus on current state-of-the-art systems for Open Information Extraction. We will also cover the NLP background required for understanding.

Topics

The topics and dates were assigned and roughly explained in the first sessions.

• NLP Basics 1/2 (POS tagging, chunking, NER), Sven Lieber
• NLP Basics 2/2 (constituent / dependency parsing), Sven Lieber

• TextRunner + KnowItAll, Haixing Gu

• Distant supervision for RE without labeled data, Toby Peoples

• OpenIE using Wikipedia, Soraya Nikousokhan

• Never Ending Language Learning (Demo / Data available), Lukas Veneziano

• ReVerb + R2A2 (Code available), Patrick Notz

• OLLIE (Code available), Max Lotstein

• PATTY (Demo available), Ezequiel Primo

• Relation Extraction With Matrix Factorization (Code available), José Licón

• TREC Question Answering Track, Jennifer Nist

• Lymba (QA System), Fabian Schillinger

• Watson (QA System) + Relation Extraction for DeepQA, Tobias Paxian

• Robust Question Answering over the Web of Linked Data (QA System), Simon Skilevic

We are no longer assigning new topics.

Sessions

Attention: Due to illness the dates might be postponed by a week or two. We will inform you about changes via email over the Daphne system (so please register), and update this webpage as soon as possible if any shifts occur.