WAC3 - 2007

Outline Program

	Fri 14 September
20:00	Welcome Cocktail
	Sat 15 September Auditorium Socrate, Place Cardinal Mercier, Louvain-la-Neuve
9:30 – 9:45	Opening Session
9:45 – 10:45	Invited Speaker: The Crúbadán Project: Corpus building for under-resourced languages Kevin Scannell, Saint Louis University, USA
10:45 – 11:15	Classifying Web corpora into domain and genre using automatic feature identification Serge Sharoff, University of Leeds, UK
	Coffee break
11:45 – 12:15	A Human Evaluation of Filtering Functions for Pattern-based Extraction of Arbitrary Relations from the Web Sebastian Blohm & Philipp Cimiano, University of Karlsruhe, Germany
12:15 – 12:45	Identification of Languages and Encodings in a Multilingual Document Anil Kumar Singh & Jagadeesh Gorla, International Institute of Information Technology, Hyderabad, India
	Lunch time
14:30 – 16:00 Cleaneval	Overview, data preparation, scoring, results of Cleaneval Marco Baroni, Francis Chantree, Adam Kilgarriff & Serge Sharoff
	Coffee break
16:30 – 18:00 Cleaneval	System descriptions
	Dimanche 16 septembre Amphithéâtre Socrate, Place Cardinal Mercier, Louvain-la-Neuve
9:00 – 10:00	Invited Speaker (to be confirmed) or Panel: "A WAC search engine"
10:00 – 10:30	CorpEus, a 'web as corpus' tool designed for the agglutinative nature of Basque Igor Leturia & Antton Gurrutxaga, Elhuyar Foundation Iñaki Alegria & Aitzol Ezeiza, University of the Basque Country, Spain
10:30 – 11:00	Implementing a BNC-Compare-able Web Corpus William Fletcher, United States Naval Academy, USA
	Coffee break
11:30 – 12:00	Yet another web crawler Fabrice Issac, Université Paris 13, France
12:00 – 12:30	textBox : a Tool for Written Corpus Linguistic Investigation Emmanuel Cartier, Université Paris 13, France
	Lunch time
14:00 – 15:30 Cleaneval	Panel: Lessons Learned, future Cleanevals
15:30 – 16:00	Closing Session

Invited speaker : Kevin Scannell

Kevin Scannell, of Saint Louis Univ., Missouri, USA, has been working with scholars of a range of smaller languages to develop web corpora for those languages : website currently lists 135 corpora/languages.

Worskshop Co-chairs