ACL-07 Tutorials Chair Report
Joakim Nivre

In response to the call for tutorial proposals, 19 proposals were 
received before the deadline of December 15, 2006. They were grouped 
into five clusters and the proposals in each cluster were reviewed 
by one international expert in addition to the tutorials chair. The 
clusters were as follows, with the number of proposals in brackets 
and the expert reviewing the proposals in each cluster following: 

Core NLP (2), Chris Manning 
Crossing borders (3), Robert Dale 
Machine learning (3), Walter Daelemans 
Corpora (3), Nancy Ide 
Dialogue (1), Diane Litman

Based on the expert opinions, the tutorials chair assisted by the 
general chair selected five proposals, taking the following criteria 
into account:

- Quality: Preference was given to proposals that were ranked highly  
  by the experts, taking both the content of the proposal and the 
  competence and experience of the presenters into account.
- Diversity: Preferably not more than one (or exceptionally two) 
  proposal(s) should be selected from each cluster. 
- Novelty: Tutorial topics featured at recent ACL events were 
  dispreferred (unless the content was clearly novel and different).
 
This resulted in the following five proposals being accepted for 
presentation:

T1 (p.m.): Bayesian Nonparametric Structured Models 
Percy Liang, Dan Klein 

T2 (a.m.): Usability and Performance Evaluation for Advanced Spoken Dialogue 
Systems 
Kristiina Jokinen, Michael McTear

T3 (p.m.): Textual Entailment 
Ido Dagan, Dan Roth, Fabio Massimo Zanzotto 

T4 (a.m.): From Web Content Mining to Natural Language Processing 
Bing Liu

T5 (a.m.): Quality Control of Corpus Annotation Through Reliability Measures 
Ron Artstein

After the deadline of notification, January 15, 2007, it was discovered 
that one additional proposal had been sent before the deadline but had 
not reached its final destination because of a temporarily malfunctioning 
server that did not relay the message properly. After discussion with the 
general chair it was decided that the proposal should be reviewed as if it 
had been received in time, since the program could still accomodate one 
additional tutorial. This review, which involved two of the international 
experts, resulted in the proposal being declined. However, the incident 
shows that e-mail without explicit acknowledgment is not a safe method 
for submission of tutorial proposals. In the future, it is therefore 
recommended that tutorial proposals are submitted through the same 
conference system as all other submissions.