Evaluation of Semantic Dependency Labeling Across Domains

AAAI Conferences

One of the key concerns in computational semantics is to construct a domain independent semantic representation which captures the richness of natural language, yet can be quickly customized to a specific domain for practical applications. We propose to use generic semantic frames defined in FrameNet, a domain-independent semantic resource, as an intermediate semantic representation for language understanding in dialog systems. In this paper we: (a) outline a novel method for FrameNet-style semantic dependency labeling that builds on a syntactic dependency parse; and (b) compare the accuracy of domain-adapted and generic approaches to semantic parsing for dialog tasks, using a frame-annotated corpus of human-computer dialogs in an airline reservation domain.


Unsupervised Induction of Semantic Roles within a Reconstruction-Error Minimization Framework

arXiv.org Machine Learning

We introduce a new approach to unsupervised estimation of feature-rich semantic role labeling models. Our model consists of two components: (1) an encoding component: a semantic role labeling model which predicts roles given a rich set of syntactic and lexical features; (2) a reconstruction component: a tensor factorization model which relies on roles to predict argument fillers. When the components are estimated jointly to minimize errors in argument reconstruction, the induced roles largely correspond to roles defined in annotated resources. Our method performs on par with most accurate role induction methods on English and German, even though, unlike these previous approaches, we do not incorporate any prior linguistic knowledge about the languages.


Leveraging Lexical Resources for the Detection of Event Relations

AAAI Conferences

Recent efforts at providing a shallow semantic analysis based on supervised systems trained on data labeled with sense tags and semantic role labels are increasingly successful (see CoNLL 2008). However, there is much more to "understanding" sentences than sense tags and semantic role labels. In particular, once the events themselves have been identified, it is essential to recognize any temporal or causal relations between them, and to infer changes in state that might have occurred (Pustejovsky et al. 2005). Resources like TimeBank are directed at the task of identifying temporal relations (Day et al. 2003), while lexical resources such as VerbNet (Dang et al. 1998) and FrameNet (Baker et al. 1998) show promise for inferring causation and enablement relations. This inference process is illustrated here with several examples. Examples are also provided of extending the range of inferences through links to an ontology such as the ISI Omega Ontology (Philpot et al. 2005).


Open Domain Information Extraction via Automatic Semantic Labeling

AAAI Conferences

This paper presents a semantic labeling technique based on information encoded in FrameNet. Sentences labeled for frames relevant to any new Information Extraction domain enable the automatic acquisition of extraction rules for the new domain. The experimental results show that both the semantic labeling and the extraction rules enabled by the labels are generated automatically with a high precision.


Sources of Complexity in Semantic Frame Parsing for Information Extraction

arXiv.org Artificial Intelligence

This paper describes a Semantic Frame parsing System based on sequence labeling methods, precisely BiLSTM models with highway connections, for performing information extraction on a corpus of French encyclopedic history texts annotated according to the Berkeley FrameNet formalism. The approach proposed in this study relies on an integrated sequence labeling model which jointly optimizes frame identification and semantic role segmentation and identification. The purpose of this study is to analyze the task complexity, to highlight the factors that make Semantic Frame parsing a difficult task and to provide detailed evaluations of the performance on different types of frames and sentences.