[Corpora-List] Reinforcement Learning from Dialogue Act

[Matthew Purver]

Dear Michael

sounds like you're thinking of the Map Task corpus. There are several 
versions now, but probably best to start here:

http://www.hcrc.ed.ac.uk/maptask/

Matt

Michael Oakes wrote:
> 
> Dear All,
> 
> As per the message below, we are looking for any question and answer 
> dialogues, where feedback is given on the quality of the answer.
> 
> I seem to remember that there is a corpus of people asking for 
> geographical directions to a person with a map, but hidden behind a 
> screen. Does anyone know any more about this?
> 
> Regards, Michael.
> 
> Martin Page wrote:
> 
>> Hi Michael.
>>
>> I'm interested in applying Reinforcement Learning to dialog acts, i 
>> was wondering if you knew of any good datasets to use. In particular 
>> I'm looking for datasets where the user asks a question about a 
>> specific question, an answer is provided and then the user provides 
>> feedback whether the information was good or bad.
>>
>> Thanks
>>
>> Martin
>>
> 
> 

-- 
Matthew Purver <mpurver at stanford.edu>
Computational Semantics Laboratory, CSLI, Stanford