Using Summarization to Discover Argument Facets in Online Idealogical Dialog
23597Authors
nan
Abstract
More and more of the information available on the web is dialogic, and a significant por- tion of it takes place in online forum conver- sations about current social and political top- ics. We aim to develop tools to summarize what these conversations are about. What are the CENTRAL PROPOSITIONS associated with different stances on an issue; what are the ab- stract objects under discussion that are central to a speaker’s argument? How can we recog- nize that two CENTRAL PROPOSITIONS real- ize the same FACET of the argument? We hy- pothesize that the CENTRAL PROPOSITIONS are exactly those arguments that people find most salient, and use human summarization as a probe for discovering them. We describe our corpus of human summaries of opinion- ated dialogs, then show how we can identify similar repeated arguments, and group them into FACETS across many discussions of a topic. We define a new task, ARGUMENT FACET SIMILARITY (AFS), and show that we can predict AFS with a .54 correlation score, versus an ngram system baseline of .39 and a semantic textual similarity system baseline of .45. 1
Keywords
Full Content
Human Language Technologies: The 2015 Annual Conference of the North American Chapter of the ACL, pages 430–440, Denver, Colorado, May 31 – June 5, 2015. c⃝2015 Association for Computational Linguistics Using Summarization to Discover Argument Facets in Online Idealogical Dialog Amita Misra, Pranav Anand, Jean Fox Tree, and Marilyn Walker UC Santa Cruz Natural Language and Dialogue Systems Lab 1156 N. High. SOE-3 Santa Cruz, California, 95064, USA amisra2|panand|foxtree|mawalker@ucsc.edu Abstract More and more of the information available on the web is dialogic, and a significant por- tion of it takes place in online forum conver- sations about current social and political top- ics. We aim to develop tools to summarize what these conversations are about. What are the CENTRAL PROPOSITIONS associated with different stances on an issue; what are the ab- stract objects under discussion that are central to a speaker’s argument? How can we recog- nize that two CENTRAL PROPOSITIONS real- ize the same FACET of the argument? We hy- pothesize that the CENTRAL PROPOSITIONS are exactly those arguments that people find most salient, and use human summarization as a probe for discovering them. We describe our corpus of human summaries of opinion- ated dialogs, then show how we can identify similar repeated arguments, and group them into FACETS across many discussions of a topic. We define a new task, ARGUMENT FACET SIMILARITY (AFS), and show that we can predict AFS with a .54 correlation score, versus an ngram system baseline of .39 and a semantic textual similarity system baseline of .45. 1 Introduction In the wake of the Penn TreeBank, much progress has been achieved in processing the monologic, informational language characteristic of newswire text. But an increasing share of the text data on the web is unlike newswire in a variety of ways: it is dialogic, opinionated, argumentative. And while some of these dialogs may be a little more than flame wars, a significant portion involve contentful, rea- PostID:Turn S1:1 Agreed She is ignoring my religious freedom and trying to institute her religion into law. The law that will bar my family from legal protections. It won’t protect her marriage but will bar me and my people from from being full citizens. She isn’t protecting marriage but perserving her heterosexual privledge. S2:1 How on earth is she impeding on you religious freedom? She isn’t trying to take away your right to any religious ceremony. With such a wide-open standard of what constitutes religious freedom that you seem to have, any legislation could be construed as imposing on religious freedom. S1:2 Because it is her religious belief that marriage is between a man and a woman. My religious belief is that marriage is between two people that love each other regardless of sex. She is tying to place her religious belif into law over mine. Who gets hurt here? If my religious belief is put into law she can still marry the person of her choice. If her religious belief gets put into law she can still marry …
Content truncated to 500 words
Labeled Text
More and more of the information available on the web is dialogic, and a significant portion of it takes place in online forum conversations about current social and political topics. We aim to develop tools to summarize what these conversations are about. What are the CENTRAL PROPOSITIONS associated with different stances on an issue; what are the abstract objects under discussion that are central to a speaker’s argument? How can we recognize that two CENTRAL PROPOSITIONS realize the same FACET of the argument? We hypothesize that the CENTRAL PROPOSITIONS are exactly those arguments that people find most salient, and use human summarization as a probe for discovering them. We describe our corpus of human summaries of opinionated dialogs, then show how we can identify similar repeated arguments, and group them into FACETS across many discussions of a topic. We define a new task, ARGUMENT FACET SIMILARITY (AFS), and show that we can predict AFS with a .54 correlation score, versus an ngram system baseline of .39 and a semantic textual similarity system baseline of .45. 1
Publication Details
- Paper ID:
- 23597
- Year:
- 2015
- Publication:
- TIPSTER
- Publication Type:
- conf
- DOI:
- nan