SUBJECT: Re : &NAME information Thanks for the argument / adjunct domain stuff -- I 'll try to rewrite my conditions ! <QUOTE> What 's the deadline for the workshop paper ? -- it might be better if I work out the non-coref conditions in &NAME terms and we treat it as a joint paper ( but I 'm quite busy for the next couple of weeks ) <END_QUOTE> The workshop deadline is tomorrow , so I think that I 'll submit what I 've got so far , which suggests that the performance of the &NAME and &NAME algorithm is on average only &NUM better when it 's using the &NAME and &NAME parser than when it uses &NAME ' GRs . ( The performance is significantly better though . ) There is a journal paper that I 'm thinking of writing : special issue for the &NAME conference . It 'd be an extension of my &NAME paper and this workshop paper ( varying full parsers in &NAME and Leass and using a shallow parser in &NAME and Leass , basic comparison with &NAME and &NAME ) . The deadline for that one is 10th March , so that would make sense to be a joint paper if you did n't mind supplying me with the necessary linguistic knowledge to make some changes to my programs ! &NAME .