max planck institut
mpii logo Minerva of the Max Planck Society

Seminar "Information Extraction & Knowledge Harvesting", Winter 2010/2011

Dr. Martin Theobald, Dr. Rainer Gemulla


Contents of the Seminar

The seminar provides an overview over information extraction (IE) techniques, the design and creation of knowledge bases, as well as their manual and automatic population. We start with a number of classic topics, such as Hearst-pattern-based IE and Wrapper Induction, before we move on to more recent trends with statistical models, declarative IE approaches, and open information extraction.

Requirements for the Certificate

Overview of Information Extraction & Knowledge Harvesting

Early Pattern-based Information Extraction

Hand-built Knowledge Representation Frameworks

Wrapper Induction

DBpedia & YAGO

Statistical Relational Learning (I): Markov Logic Networks & Applications

Wikipedia-based Information Extraction Frameworks (Kylin/KOG)

Segmentation & Disambiguation

Declarative IE: SystemT

Open IE

Iterative IE & Provenance