When: 2013-02-06 at 09:00-10:00
Presenters: Peter Exner, Pierre Nugues
Where: E:2116, E-house, LTH
Abstract: Text is the largest repository of human knowledge and many of the most popular web applications today, Google, Yahoo, or Twitter, are based on text. While there are many ways to model, extract, and process knowledge in text, semantic models are key elements to many advanced applications in text processing. Semantic interpretation is then instrumental to automated knowledge processing. This includes information extraction, machine translation, summarization, and question answering.
In this LUCAS seminar, we will describe our on-going work in semantic processing for English, Swedish, and other languages that includes the conversion of text into animated 3D scenes, the design of multilingual generic semantic parsers and coreference solvers, the extraction of propositions and entities from large corpora (Wikipedia). We will also address scalability issues and our use of cluster-based software relying on Hadoop.