Cross-topic Argument Mining from Heterogeneous Sources

By Christian Stab, Tristan Miller, Benjamin Schiller, Pranav Rai, Iryna Gurevych. 2018

219 citations recorded by the source corpus.

Argument mining is a core technology for automating argument search in large document collections. Despite its usefulness for this task, most current approaches are designed for use only with specific text types and fall short when applied to heterogeneous texts. In this paper, we propose a new sentential annotation scheme that is reliably applicable by crowd workers to arbitrary Web texts. We source annotations for over 25,000 instances covering eight controversial topics. We show that integrating topic information into bidirectional long short-term memory networks outperforms vanilla BiLSTMs by more than 3 percentage points in F1 in two- and three-label cross-topic settings. We also show that these results can be further improved by leveraging additional data for topic relevance using multi-task learning.

Subfields: Artificial Intelligence, Information Systems, Natural Language Processing, Information Extraction

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

In Information Systems

In Natural Language Processing

In Information Extraction

Citations in the corpus

AdapterFusion: Non-Destructive Task Composition for Transfer Learning (2020, methodology): “7 The results for Commonsense QA and ArgumentMining are in the Appendix, which is consistent with our ﬁndings on SST-2. learning rate of 1 e − 4 .”
An Empirical Investigation of the Role of Pre-training in Lifelong Learning (2021, background): “The Argument Aspect Corpus ( AAC ) (Stab et al., 2018) has over 25,000 arguments spanning eight topics annotated with three labels (no argument, supporting argument, opposing argument).”
Let’s Make Your Request More Persuasive: Modeling Persuasive Strategies via Semi-Supervised Neural Nets on Crowdfunding Platforms (2019, background): “Computational argumentation has received much recent attention (Ghosh et al., 2016; Stab and Gurevych, 2017; Peldszus and Stede, 2013; Stab et al., 2018; Ghosh et al., 2014).”
TU Wien at TREC DL and Podcast 2021: Simple Compression for Dense Retrieval (2021, methodology): “We utilize RoBERTArg 4 , which is trained on an argument/non-argument labeled dataset [22].”
Augmented SBERT: Data Augmentation Method for Improving Bi-Encoders for Pairwise Sentence Scoring Tasks (2020, unknown): “We use the dataset by Stab et al. (2018), which contains pro and con stance arguments for eight controversial topics (T1 - T8) (“cloning”, “abortion”, “minimum wage”, “marijuana legalization”, “nuclear energy”, “death penalty”, “gun control”, “school uniforms”) retrieved from heterogeneous web…”
Corpus Wide Argument Mining - a Working Solution (2019, methodology): “Similarly, Rinott et al. (2015) aimed to detect evidence for 58 different topics, over 547 manually-collected articles, and Stab et al. (2018) did so for 8 controversial topics, while considering the top 50 Google-retrieved web pages for each topic.”
Aspect-Controlled Neural Argument Generation (2020, methodology): “We use the ArgumenText API’s 5 argument and stance classiﬁ-cation models (Stab et al., 2018a) to classify all sentences into arguments or non-arguments (F 1 macro = .”