From TreeBank to PropBank
By Paul R. Kingsbury, Martha Palmer. 2002
785 citations recorded by the source corpus.
This paper describes our approach to the development of a Proposition Bank, which involves the addition of semantic information to the Penn English Treebank. Our primary goal is the labeling of syntactic nodes with specific argument labels that preserve the similarity of roles such as the window in John broke the window and the window broke . After motivating the need for explicit predicate argument structure labels, we briefly discuss the theoretical considerations of predicate argument structure and the need to maintain consistency across syntactic alternations. The issues of consistency of argument structure across both polysemous and synonymous verbs are also discussed and we present our actual guidelines for these types of phenomena, along with numerous examples of tagged sentences and verb frames. Metaframes are introduced as a technique for handling similar frames among near− synonymous verbs. We conclude with a summary of the current status of annotation process.
Subfields: Artificial Intelligence, Computational Linguistics & Semantics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Luke S. Zettlemoyer
- Mohit Bansal
- Noah A. Smith
- Christopher D. Manning
- Percy Liang
- Yejin Choi
- Jianfeng Gao
- Hannaneh Hajishirzi
In Computational Linguistics & Semantics
- Alessandro Moschitti
- Dan Roth
- Alessandro Sperduti
- Davide Bacciu
- Alessio Micheli
- Wen-tau Yih
- Noah A. Smith
- Ralph Grishman
Citations in the corpus
- A semantic matching energy function for learning with multi-relational data (2013, background): “1998) and by Giuglea and Moschitti (2006), who proposed a model to connect WordNet, VerbNet and PropBank (Kingsbury and Palmer 2002) for semantic parsing using tree kernels.”
- Question-Answer Driven Semantic Role Labeling: Using Natural Language to Annotate Natural Language (2015, background): “PropBank (Kingsbury and Palmer, 2002), NomBank (Meyers et al., 2004) and OntoNotes (Hovy et al., 2006) circumvent the need for a large lexicon of roles, by defin- 3Our hope is that this approach will generalize not only across different domains in English, as we show in this paper, but also to…”
- “Going on a vacation” takes longer than “Going for a walk”: A Study of Temporal Commonsense Understanding (2019, methodology): “Therefore, to perturb such candidates, we create a pool of 60 k event phrases using PropBank (Kingsbury and Palmer, 2002), and perturb the candidate answers to be the most similar ones extracted by an information retrieval ( IR ) system.”
- Automatically Solving Number Word Problems by Semantic Parsing and Reasoning (2015, methodology): “Our future plan is automatically generating verb functions from databases like PropBank (Kingsbury & Palmer, 2002), FrameNet (Fillmore et al., 2003), and VerbNet 4 (Schuler, 2005).”
- Question Answering as Global Reasoning Over Semantic Abstractions (2018, background): “Many SRL systems have been designed (Gildea and Jurafsky 2002; Punyakanok, Roth, and Yih 2008) using linguistic resources such as FrameNet (Baker, Fillmore, and Lowe 1998), PropBank (Kingsbury and Palmer 2002), and NomBank (Meyers et al.”
- PPDB: The Paraphrase Database (2013, methodology): “To estimate the usefulness of PPDB as a resource for tasks like semantic role labeling or parsing, we analyze its coverage of Propbank predicates and predicate-argument tuples (Kingsbury and Palmer, 2002).”
- The Necessity of Parsing for Predicate Argument Recognition (2002, methodology): “The Propbank project at the University of Pennsylvania (Kingsbury and Palmer, 2002) and the FrameNet project at the International Computer Science Institute (Baker et al., 1998) share the goal of documenting the syntactic realization of arguments of the predicates of the general English lexicon by a”
- The Penn Discourse Treebank (2004, background): “This level is deeper than that provided by PTB and Propbank, yet it is shallow in the sense that it peels off just those inferences licensed by the connectives.”
- A large-scale classification of English verbs (2008, methodology): “The coverage of VN 1.0 has been evaluated through a mapping to almost 50,000 instances from Proposition Bank's corpus instances (Kingsbury and Palmer, 2002).”
- Efficient Convolution Kernels for Dependency and Constituent Syntactic Trees (2006, methodology): “We experimented with such kernels and Support Vector Machines (SVMs) on (a) the classification of semantic roles defined in PropBank [8] and FrameNet [9] and (b) the classification of questions from Question Answering scenarios.”
- Extracting Relations with Integrated Information Using Kernel Methods (2005, methodology): “Such analysis may be based on newly-annotated corpora like PropBank (Kingsbury and Palmer, 2002) at the University of Pennsylvania and NomBank (Meyers et al., 2004) at New York University.”
- The NomBank Project: An Interim Report (2004, methodology): “PropBank (Kingsbury et al., 2002; Kingsbury and Palmer, 2002; University of Pennsylvania, 2002), Nom-Bank and other annotation projects taken together should lead to the creation of better tools for the automatic analysis of text.”
- Calibrating Features for Semantic Role Labeling (2004, background): “The availability of semantically annotated corpora such as the Proposition Banks (Kingsbury and Palmer, 2002; Xue and Palmer, 2003) and FrameNet (Baker et al., 1998) have enabled the development of a rapidly growing list of statistical semantic analyzers (Gildea and Jurafsky, 2002; Gildea and…”
- The Importance of Syntactic Parsing and Inference in Semantic Role Labeling (2008, methodology): “For more information on PropBank and the semantic role labeling task, readers can refer to (Kingsbury and Palmer, 2002) and (Carreras and Màrquez, 2004; Carreras and Màrquez, 2005).”
- Shallow Semantic Parsing Based on FrameNet, VerbNet and PropBank (2006, methodology): “On FrameNet, we obtain the gold Levin class annotation through a mapping between Intersective Levin classes and FrameNet frames [7].”
- Syntactic and Semantic Kernels for Short Text Pair Categorization (2009, methodology): “In particular, in (Moschitti et al., 2007) kernels for the processing of PASs (in PropBank 1 format (Kingsbury and Palmer, 2002)) extracted from question/answer pairs were proposed.”
- Generative Kernels for Tree-Structured Data (2018, methodology): “We employ a version of the Propbank data set [49] introduced in [50], that includes a TABLE III”
- Identifying Relations for Open Information Extraction (2011, neutral): “Finally, SRL requires hand-constructed semantic resources like Propbank and Framenet (Martha and Palmer, 2002; Baker et al., 1998) as input.”