... problem (Scott and Moore,2007): Task -based evaluations with human experi-mental subjects are time-consuming and expensive, and corpus -based evaluations of NLG systems areproblematic because a ... 374vs.n = 91), and offers other oppor-tunities for more fine-grained analysis as well. Wetake this as an empirical validation of the Internet- based evaluation of GIVE, and propose that it ... Proceedings of the ACL-IJCNLP 2009 Conference Short Papers, pages 301–304,Suntec, Singapore, 4 August 2009.c2009 ACL and AFNLPValidating the web -based evaluation of NLG systems Alexander KollerSaarland...