Existing work in the extraction of commonsense knowledge from text has been primarily restricted to factoids that serve as statements about what may possibly obtain in the world. We present an approach to deriving stronger, more general claims by abstracting over large sets of factoids. Our goal is to coalesce the observed nominals for a given predicate argument into a few predominant types, obtained as WordNet synsets. The results can be construed as generically quantified sentences restricting the semantic type of an argument position of a predicate. 1

Supervised semantic role labeling (SRL) systems are generally claimed to have accuracies in the range of 80 % and higher (Erk and Padó, 2006). These numbers, though, are the result of highly-restricted evaluations, i.e., typically evaluating on hand-picked lemmas for which training data is available. In this paper we consider performance of such systems when we evaluate at the document level rather than on the lemma level. While it is wellknown that coverage gaps exist in the resources available for training supervised SRL systems, what we have been lacking until now is an understanding of the precise nature of this coverage problem and its impact on the performance of SRL systems. We present a typology of five different types of coverage gaps in FrameNet. We then analyze the impact of the coverage gaps on performance of a supervised semantic role labeling system on full texts, showing an average oracle upper bound of 46.8%.

Many applications in the context of natural language processing or information retrieval may be largely improved if they were able to fully exploit the rich semantic information annotated in high-quality, publicly available resources such as the FrameNet and the Word-Net databases. Nevertheless, the practical use of similar resources is often biased by the limited coverage of semantic phenomena that they provide. A natural solution to this problem would be to automatically establish anchors between these resources that would allow us 1) to jointly use the encoded information, thus possibly overcoming limitations of the individual corpora, and 2) to extend each resource coverage by exploiting the information encoded in the others. In this paper, we present a supervised learning framework for the mapping of FrameNet lexical units onto WordNet synsets based on a reduced set of novel and semantically rich features. The automatically learnt mapping, which we call MapNet, can be used 1) to extend frame sets in the English FrameNet, 2) to populate frame sets in the Italian FrameNet via MultiWordNet and 3) to add frame labels to the MultiSemCor corpus. Our evaluation on these tasks shows that the proposed approach is viable and can result in accurate automatic annotations. 1

annual meetings organized by SIGNLL, the ACL special interest group on natural language learning. CoNLL-2009 will be held in Boulder, CO, 4–5 June 2009, in conjunction with NAACL HLT. For our special focus this year in the main session of CoNLL, we invited papers on unsupervised,

novel approach to mapping

This paper focuses on the improvement of the conceptual structure of FrameNet for the sake of applying this resource to knowledgeintensive NLP tasks requiring reasoning, such as question answering, information extraction etc. Ontological analysis supported by data-driven methods is used for axiomatizing, enriching and cleaning up frame relations. The impact of the achieved axiomatization is investigated on recognizing textual entailment. 1.