In this paper we describe an unsupervised method for semantic role induction which holds promise for relieving the data acquisition bottleneck associated with supervised role labelers .

By combining role induction with a rule-based component for argument identification we obtain an unsupervised end-to-end semantic role labeling system.

Coupled with a rule-based component for automatically identifying argument candidates our split-merge algorithm forms an end-to-end system that is capable of inducing role labels without any supervision.

Although the dataset provides annotations for verbal and nominal predicate-argument constructions, we only considered the former, following previous work on semantic role labeling (Marquez et al., 2008).

This baseline has been previously used as point of comparison by other unsupervised semantic role labeling systems (Grenager and Manning, 2006; Lang and Lapata, 2010) and shown difficult to outperform.

Indeed, the analysis produced by existing semantic role labelers has been shown to benefit a wide spectrum of applications ranging from information extraction (Surdeanu et al., 2003) and question answering (Shen and Lapata, 2007), to machine translation (Wu and Fung, 2009) and summarization (Melli et al., 2005).

mantic role labeling as a supervised learning problem.

Unfortunately, the reliance on role-annotated data which is expensive and time-consuming to produce for every language and domain, presents a major bottleneck to the widespread application of semantic role labeling .

We follow the general architecture of supervised semantic role labeling systems.

Swier and Stevenson (2004) induce role labels with a bootstrapping scheme where the set of labeled instances is iteratively expanded using a classifier trained on previously labeled instances.

We present the results of evaluating translation utility by measuring the accuracy within a semantic role labeling (SRL) framework.

Finally, we show that replacing the human semantic role labelers with an automatic shallow semantic parser in our proposed metric yields an approximation that is about 80% as closely correlated with human judgment as HTER, at an even lower cost—and is still far better correlated than n-gram based evaluation metrics.

Table 7: Inter-annotator agreement rate on role classification (matching of role label associated with matched word span)

Role labels (A0, MTMP, etc.)

Before annotation began, all semantic information was removed by mapping all role labels to ARG.

Because PropBank adds semantic role annotation on top of the Penn TreeB ank, we have available syntactic annotation and semantic role labels for all instances.

State-of-the-art semantic role labelers (e.g., the ones trained over PropBank) do not completely represent the meaning of negated statements.

For all statements s, current role labelers would only encode it is not the case that s. However, examples (1—7)