From Surrogacy to Adoption; From Bitcoin to Cryptocurrency: Debate Topic Expansion

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 977–990 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 977 From Surrogacy to Adoption; From Bitcoin to Cryptocurrency: Debate Topic Expansion Roy Bar-Haim∗, Dalia Krieger∗, Orith Toledo-Ronen∗, Lilach Edelstein, Yonatan Bilu, Alon Halfon, Yoav Katz, Amir Menczel, Ranit Aharonov and Noam Slonim IBM Research Abstract When debating a controversial topic, it is of- ten desirable to expand the boundaries of dis- cussion. For example, we may consider the pros and cons of possible alternatives to the de- bate topic, make generalizations, or give spe- cific examples. We introduce the task of De- bate Topic Expansion - finding such related topics for a given debate topic, along with a novel annotated dataset for the task. We fo- cus on relations between Wikipedia concepts, and show that they differ from well-studied lexical-semantic relations such as hypernyms, hyponyms and antonyms. We present algo- rithms for finding both consistent and con- trastive expansions and demonstrate their ef- fectiveness empirically. We suggest that de- bate topic expansion may have various use cases in argumentation mining. 1 Introduction Recent years saw substantial advancement of De- bating Technologies – computational technologies developed directly to enhance, support, and en- gage with human debating (Gurevych et al., 2016). A recent milestone in this field is IBM R⃝’s Project Debater R⃝1, the first demonstration of a live com- petitive debate between an AI system and a human debate champion. When debating a controversial topic, it is often desirable to expand the boundaries of the discus- sion, and bring up arguments about related topics. For example, when discussing the pros and cons of the presidential system, it is natural to contrast it with those of the parliamentary system. When de- bating alternative medicine, we may discuss spe- cific examples, such as homeopathy and naturopa- thy. Conversely, when discussing bitcoins, we can speak more broadly on cryptocurrency. ∗First three authors equally contributed to this work. 1https://www.research.ibm.com/artificial- intelligence/project-debater/ Consider the use of debating technologies for decision support, where the pros and cons of a given proposal are extracted from a large cor- pus, summarized and presented to the user. Cur- rent methods for topic-related, corpus-wide argu- ment mining only specify the given debate topic in their search queries (Levy et al., 2017, 2018; Wachsmuth et al., 2017; Stab et al., 2018a,b). As a result, much of the relevant argumentative content is left out of their reach. Alternatively, context-independent argument mining can exhaus- tively extract argumentative content from a cor- pus (Lippi and Torroni, 2015), but it cannot tell which arguments are actually relevant for the topic in question. In this work we take a step towards closing this gap, by introducing the task of Debate Topic Ex- pansion – finding related topics that can enrich our arguments and strengthen our case when debat- ing a given topic. Following previous work (Levy et al., 2017, 2018), we focus on topics that are Wikipedia concepts (article titles …

Content truncated to 500 words