Recognising Agreement and Disagreement between Stances with Reason Comparing Networks

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4665–4671 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 4665 Recognising Agreement and Disagreement between Stances with Reason Comparing Networks Chang Xu, C´ecile Paris, Surya Nepal, and Ross Sparks CSIRO Data61, Australia {Chang.Xu, Cecile.Paris, Surya.Nepal, Ross.Sparks}@data61.csiro.au Abstract We identify agreement and disagreement be- tween utterances that express stances towards a topic of discussion. Existing methods fo- cus mainly on conversational settings, where dialogic features are used for (dis)agreement inference. We extend this scope and seek to detect stance (dis)agreement in a broader setting, where independent stance-bearing ut- terances, which prevail in many stance cor- pora and real-world scenarios, are compared. To cope with such non-dialogic utterances, we find that the reasons uttered to back up a specific stance can help predict stance (dis)agreements. We propose a reason com- paring network (RCN) to leverage reason in- formation for stance comparison. Empirical results on a well-known stance corpus show that our method can discover useful reason in- formation, enabling it to outperform several baselines in stance (dis)agreement detection. 1 Introduction Agreement and disagreement naturally arise when peoples’ views, or “stances”, on the same topics are exchanged. Being able to identify the con- vergence and divergence of stances is valuable to various downstream applications, such as discov- ering subgroups in a discussion (Hassan et al., 2012; Abu-Jbara et al., 2012), improving recog- nition of argumentative structure (Lippi and Tor- roni, 2016), and bootstrapping stance classifica- tion with (dis)agreement side information (Sridhar et al., 2014; Ebrahimi et al., 2016). Previous efforts on (dis)agreement detection are confined to the scenario of natural dia- logues (Misra and Walker, 2013; Wang and Cardie, 2014; Sridhar et al., 2015; Rosenthal and McKeown, 2015), where dialogic struc- tures are used to create a conversational con- text for (dis)agreement inference. However, non- dialogic stance-bearing utterances are also very common in real-world scenarios. For example, in social media, people can express stances au- tonomously, without the intention of initiating a discussion (Mohammad et al., 2016). There are also corpora built with articles containing many self-contained stance-bearing utterances (Ferreira and Vlachos, 2016; Bar-Haim et al., 2017). Studying how to detect (dis)agreement between such independent stance-bearing utterances has several benefits: 1) pairing these utterances can lead to a larger (dis)agreement corpus for train- ing a potentially richer model for (dis)agreement detection; 2) the obtained pairs enable train- ing a distance-based model for opinion cluster- ing and subgroup mining; 3) it is applicable to the aforementioned non-dialogic stance corpora; and 4) it encourages discovering useful signals for (dis)agreement detection beyond dialogic features (e.g., the reason information studied in this work). In this work, we investigate how to detect (dis)agreement between a given pair of (presum- ably unrelated) stance-bearing utterances. Table 1 shows an example where a decision is made on whether two utterances agree or disagree on a dis- cussed topic. This task, however, is more chal- lenging, as the inference has to be made without …

Content truncated to 500 words