Almost all automatic semantic role labeling (SRL) systems rely on a preliminary parsing step that derives a syntactic structure from the sentence being analyzed. This makes the choice of syntactic representation an essential design decision. In this paper, we study the influence of syntactic representation on the performance of SRL systems. Specifically, we compare constituent-based and dependencybased representations for SRL of English in the FrameNet paradigm. Contrary to previous claims, our results demonstrate that the systems based on dependencies perform roughly as well as those based on constituents: For the argument classification task, dependencybased systems perform slightly higher on average, while the opposite holds for the argument identification task. This is remarkable because dependency parsers are still in their infancy while constituent parsing is more mature. Furthermore, the results show that dependency-based semantic role classifiers rely less on lexicalized features, which makes them more robust to domain changes and makes them learn more efficiently with respect to the amount of training data.

The SemEval-2007 task to disambiguate prepositions was designed as a lexical sample task. A set of over 25,000 instances was developed, covering 34 of the most frequent English prepositions, with two-thirds of the instances for training and one-third as the test set. Each instance identified a preposition to be tagged in a full sentence taken from the FrameNet corpus (mostly from the British National Corpus). Definitions from the Oxford Dictionary of English formed the sense inventories. Three teams participated, with all achieving supervised results significantly better than baselines, with a high fine-grained precision of 0.693. This level is somewhat similar to results on lexical sample tasks with open class words, indicating that significant progress has been made. The data generated in the task provides ample opportunitites for further investigations of preposition behavior. 1

We present the psycholinguistically motivated task of predicting human plausibility judgements for verb-role-argument triples and introduce a probabilistic model that solves it. We also evaluate our model on the related role-labelling task, and compare it with a standard role labeller. For both tasks, our model benefits from classbased smoothing, which allows it to make correct argument-specific predictions despite a severe sparse data problem. The standard labeller suffers from sparse data and a strong reliance on syntactic cues, especially in the prediction task.

This article describes how semantic role resources can be exploited for preposition disambiguation. The main resources include the semantic role annotations provided by the Penn Treebank and FrameNet tagged corpora. The resources also include the assertions contained in the Factotum knowledge base, as well as information from Cyc and Conceptual Graphs. A common inventory is derived from these in support of definition analysis, which is the motivation for this work. The disambiguation concentrates on relations indicated by prepositional phrases, and is framed as word-sense disambiguation for the preposition in question. A new type of feature for word-sense disambiguation is introduced, using WordNet hypernyms as collocations rather than just words. Various experiments over the Penn Treebank and FrameNet data are presented, including prepositions classified separately versus together, and illustrating the effects of filtering. Similar experimentation is done over the Factotum data, including a method for inferring likely preposition usage from corpora, as knowledge bases do not generally indicate how relationships are expressed in English (in contrast to the explicit annotations on this in the Penn Treebank and FrameNet). Other experiments are included with the FrameNet data mapped into the common relation inventory developed for definition analysis, illustrating how preposition disambiguation might be applied in lexical acquisition. 1.

To answer the critical need for sharable, reusable annotated resources with rich linguistic annotations, we are developing a Manually Annotated Sub-Corpus (MASC) including texts from diverse genres and manual annotations or manually-validated annotations for multiple levels, including WordNet senses and FrameNet frames and frame elements, both of which have become significant resources in the international computational linguistics community. To derive maximal benefit from the semantic information provided by these resources, the MASC will also include manually-validated shallow parses and named entities, which will enable linking WordNet senses and FrameNet frames within the same sentences into more complex semantic structures and, because named entities will often be the role fillers of FrameNet frames, enrich the semantic and pragmatic information derivable from the sub-corpus. All MASC annotations will be published with detailed inter-annotator agreement measures. The MASC and its annotations will be freely downloadable from the ANC website, thus providing maximum accessibility for researchers from around the globe. 1. Overview To answer the critical need for sharable, reusable annotated resources with rich linguistic annotations, we are developing a Manually Annotated Sub-Corpus

Abstract. This paper describes a new approach for describing contents through the use of interlinguas in order to facilitate the extraction of specific pieces of information. The authors highlight the different dimensions of a document and how these dimensions define the capacities of their respective contents to be found in the scalable process of finding information. A specific interlingua, UNL, will be described. This approach is illustrated both with rich examples of the followed model and with actual applications, that includes the description of some running projects based on the interlingual representation of contents.

The encoding of selectional constraints refers to such features. Chomsky claims that they are syntactic features, because they play a role in purely syntactic rules. For example, the sentence (2.55) *The book who you read was a best seller.

We describe a preliminary version of Mutaphrase, a system that generates paraphrases of semantically labeled input sentences using the semantics and syntax encoded in FrameNet, a freely available lexicosemantic database. The algorithm generates a large number of paraphrases with a wide range of syntactic and semantic distances from the input. For example, given the input “I like eating cheese”, the system outputs the syntactically distant “Eating cheese is liked by me”, the semantically distant “I fear sipping juice”, and thousands of other sentences. The wide range of generated paraphrases makes the algorithm ideal for a range of statistical machine learning problems such as machine translation and language modeling as well as other semanticsdependent tasks such as query and language generation. 1

This paper describes our attempt at NomBank-based automatic Semantic Role Labeling (SRL). NomBank is a project at New York University to annotate the argument structures for common nouns in the Penn Treebank II corpus. We treat the NomBank SRL task as a classification problem and explore the possibility of adapting features previously shown useful in PropBank-based SRL systems. Various NomBank-specific features are explored. On test section 23, our best system achieves F1 score of 72.73 (69.14) when correct (automatic) syntactic parse trees are used. To our knowledge, this is the first reported automatic NomBank SRL system.

This task consists of recognizing words and phrases that evoke semantic frames as defined in the FrameNet project