The Penn Discourse TreeBank 2.0.

By Rashmi Prasad, N. Dinesh, Alan Lee, Eleni Miltsakaki, Livio Robaldo, Aravind K. Joshi, Bonnie Lynn Webber. International Conference on Language Resources and Evaluation, 2008

1,454 citations recorded by the source corpus.

We present the second version of the Penn Discourse Treebank, PDTB-2.0, describing its lexically-grounded annotations of discourse relations and their two abstract object arguments over the 1 million word Wall Street Journal corpus. We describe all aspects of the annotation, including (a) the argument structure of discourse relations, (b) the sense annotation of the relations, and (c) the attribution of discourse relations and each of their arguments. We list the differences between PDTB-1.0 and PDTB-2.0. We present representative statistics for several aspects of the annotation in the corpus.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/a6c26e574d74346a2b5ade5412aee2138075463e

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

1. Yejin Choi
2. Mohit Bansal
3. Percy Liang
4. Jianfeng Gao
5. Noah A. Smith
6. Luke S. Zettlemoyer
7. Graham Neubig
8. Iryna Gurevych

Citations in the corpus

• DisSent: Learning Sentence Representations from Explicit Discourse Relations (2019, methodology): “We further evaluate both the DisSent model and a BERT model fine-tuned on DisSent on two clas-sification tasks from the Penn Discourse Treebank (PDTB) (Rashmi et al., 2008).”
• Argumentation Mining on the Web from Information Seeking Perspective (2014, background): “(2013a) select five argumentation schemes from Walton and map these patterns to discourse relation categories in the Penn Discourse TreeBank (PDTB) (Prasad et al., 2008), but later they define two new schemes that they discovered in PDTB.”
• Recent advances in deep learning based dialogue systems: a systematic survey (2021, methodology): “Open-domain Dialogue Prasad et al. (2008) Luna Dialogue dataset with Italian rela-tion annotations.”
• A large-scale comparison of human-written versus ChatGPT-generated essays (2023, positive): “In the present paper, we employ the PDTB list of discourse markers 55 which we adjust to exclude words that are often used for purposes other than indicating discourse relations, such as ‘like’, ‘for’, ‘in’ etc.”
• Identifying Argumentative Discourse Structures in Persuasive Essays (2014, background): “The PDTB includes implicit as well as explicit discourse relations of different types, and there are multiple approaches aiming at automatically identifying implicit relations.”
• Implicit Discourse Relation Identification for Open-domain Dialogues (2019, background): “Our pipeline targets the four level-1 discourse relations, i.e., “Comparison”, “Expansion”, “Contin-gency” and “Temporal”.”
• A PDTB-styled end-to-end discourse parser (2012, methodology): “One possibility is to follow Prasad et al. (2010) to use a set of filters and heuristics to locate the positions of Arg1 spans.”
• Machine Comprehension with Discourse Relations (2015, background): “The set of these relations does not capture the richness of discourse representations considered in traditional discourse theories (Mann and Thompson, 1988; Prasad et al., 2008).”
• StaQC: A Systematically Mined Question-Code Dataset from Stack Overflow (2018, methodology): “We used the common connective words and phrases from Penn Discourse Tree Bank [43].”
• DisSent: Sentence Representation Learning from Explicit Discourse Relations (2017, methodology): “, 2003) or compiling a comprehensive set of discourse relations (Rashmi et al., 2008; Hobbs, 1979, 1985; Jasinskaja and Karagjosova, 2015; Knott, 1996).”
• Mining Discourse Markers for Unsupervised Sentence Representation Learning (2019, methodology): “Extensive research in linguistics has resulted in elaborate discourse marker inventories for many languages.2 These inventories were created by manual corpus exploration or annotation of small-scale corpora: the largest annotated corpus, the English PDTB consists of a few tens of thousand examples, ”
• Equal But Not The Same: Understanding the Implicit Relationship Between Persuasive Images and Text (2018, background): “In this light, our problem of determining the relationship between image and text bears some resemblance to the problem of automatically determining discourse relations between sentences, for which there is extensive prior work [2, 4, 28, 29, 30, 31].”
• Revisiting Readability: A Unified Framework for Predicting Text Quality (2008, background): “The Penn Discourse Treebank (Prasad et al., 2008) is a new resource with annotations of discourse connectives and their senses in the Wall Street Journal portion of the Penn Treebank (Marcus et al.”
• Guided Generation of Cause and Effect (2020, background): “More related are PDTB [Prasad et al. , 2008] and BECauSE [Dunietz et al. , 2017b], but where our resource goal is a much larger corpus, for the purpose of training a neural text generation model.”
• NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentation (2021, methodology): “This transformation uses the corpus analysis on PDTB 2.0 (Prasad et al., 2008) to identify discourse markers that are associated with a discourse relation with a chance of at least 0.5.”