Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


AGokcen-MA-Thesis.pdf (481.74 KB)

ETD Abstract Container

Abstract Header

A Matter of Debate: Using Dialogue Relation Labels to Augment (Dis)agreement Analysis of Debate Data

Gokcen, Ajda Zeynep
http://orcid.org/0000-0003-0347-727X
http://rave.ohiolink.edu/etdc/view?acc_num=osu1462813013

Abstract Details

2016, Master of Arts, Ohio State University, Linguistics.
Some of the most linguistically rich sources of natural conversational data come from online discussions, and from debate forums in particular. Yet existing accounts of dialogue are still largely tailored to spoken conversation, making analysis of such written, argumentative data difficult. In this thesis, I attempt to more effectively model human intuitions of (dis)agreement for a corpus of online debates. I accomplish this by developing a taxonomy of dialogue labels to accommodate the uniqueness of online debate data, and by focusing on a theoretically-grounded account of discourse acceptance and rejection as they can be found in written natural language. These labels are then used to manually annotate the debate corpus; the annotations are used to train a statistical model that classifies the level of (dis)agreement between participants in the online debates. The models successfully recreate human intuitions of (dis)agreement, achieving an accuracy of 89% for a two-way classifier between agreement and disagreement and an accuracy of 70% for a three-way classifier that includes a neutral category.
Marie-Catherine de Marneffe (Advisor)
Brian Joseph (Committee Member)
84 p.
Linguistics
linguistics; pragmatics; discourse analysis

Recommended Citations

Citations

  • Gokcen, A. Z. (2016). A Matter of Debate: Using Dialogue Relation Labels to Augment (Dis)agreement Analysis of Debate Data [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1462813013

    APA Style (7th edition)

  • Gokcen, Ajda. A Matter of Debate: Using Dialogue Relation Labels to Augment (Dis)agreement Analysis of Debate Data. 2016. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1462813013.

    MLA Style (8th edition)

  • Gokcen, Ajda. "A Matter of Debate: Using Dialogue Relation Labels to Augment (Dis)agreement Analysis of Debate Data." Master's thesis, Ohio State University, 2016. http://rave.ohiolink.edu/etdc/view?acc_num=osu1462813013

    Chicago Manual of Style (17th edition)

osu1462813013
674
© 2016, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.