PolArg: Unsupervised Polarity Prediction of Arguments in Real-Time Online Conversations. Lenz, M. & Bergmann, R. In Cimiano, P., Frank, A., Kohlhase, M., & Stein, B., editors, Robust Argumentation Machines, volume 14638, of Lecture Notes in Computer Science, pages 108–126, Cham, 2024. Springer Nature Switzerland. ![PolArg: Unsupervised Polarity Prediction of Arguments in Real-Time Online Conversations [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper doi abstract bibtex The increasing usage of social networks has led to a growing number of discussions on the Internet that are a valuable source of argumentation that occurs in real time. Such conversations are often made up of a large number of participants and are characterized by a fast pace. Platforms like X/Twitter and Hacker News (HN) allow users to respond to other users' posts, leading to a tree-like structure. Previous work focused on training supervised models on datasets obtained from debate portals like Kialo where authors provide polarity labels (i.e., support/attack) together with their posts. Such classifiers may yield suboptimal predictions for the noisier posts from X or HN, so we propose unsupervised prompting strategies for large language models instead. Our experimental evaluation found this approach to be more effective for X conversations than a model fine-tuned on Kialo debates, but less effective for HN posts (which are more technical and less argumentative). Finally, we provide an open-source application for converting discussions on these platforms into argument graphs.
@inproceedings{Lenz2024PolArgUnsupervisedPolarity,
title = {{{PolArg}}: {{Unsupervised Polarity Prediction}} of~{{Arguments}} in~{{Real-Time Online Conversations}}},
shorttitle = {{{PolArg}}},
booktitle = {Robust {{Argumentation Machines}}},
author = {Lenz, Mirko and Bergmann, Ralph},
editor = {Cimiano, Philipp and Frank, Anette and Kohlhase, Michael and Stein, Benno},
year = {2024},
series = {Lecture {{Notes}} in {{Computer Science}}},
volume = {14638},
pages = {108--126},
publisher = {Springer Nature Switzerland},
address = {Cham},
doi = {10.1007/978-3-031-63536-6_7},
abstract = {The increasing usage of social networks has led to a growing number of discussions on the Internet that are a valuable source of argumentation that occurs in real time. Such conversations are often made up of a large number of participants and are characterized by a fast pace. Platforms like X/Twitter and Hacker News (HN) allow users to respond to other users' posts, leading to a tree-like structure. Previous work focused on training supervised models on datasets obtained from debate portals like Kialo where authors provide polarity labels (i.e., support/attack) together with their posts. Such classifiers may yield suboptimal predictions for the noisier posts from X or HN, so we propose unsupervised prompting strategies for large language models instead. Our experimental evaluation found this approach to be more effective for X conversations than a model fine-tuned on Kialo debates, but less effective for HN posts (which are more technical and less argumentative). Finally, we provide an open-source application for converting discussions on these platforms into argument graphs.},
isbn = {978-3-031-63536-6},
langid = {english},
url = {https://www.wi2.uni-trier.de/shared/publications/Lenz2024PolArgUnsupervisedPolarity.pdf}
}
Downloads: 0
{"_id":"qxNC7JEqk2yWdZMSP","bibbaseid":"lenz-bergmann-polargunsupervisedpolaritypredictionofargumentsinrealtimeonlineconversations-2024","author_short":["Lenz, M.","Bergmann, R."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"PolArg: Unsupervised Polarity Prediction of Arguments in Real-Time Online Conversations","shorttitle":"PolArg","booktitle":"Robust Argumentation Machines","author":[{"propositions":[],"lastnames":["Lenz"],"firstnames":["Mirko"],"suffixes":[]},{"propositions":[],"lastnames":["Bergmann"],"firstnames":["Ralph"],"suffixes":[]}],"editor":[{"propositions":[],"lastnames":["Cimiano"],"firstnames":["Philipp"],"suffixes":[]},{"propositions":[],"lastnames":["Frank"],"firstnames":["Anette"],"suffixes":[]},{"propositions":[],"lastnames":["Kohlhase"],"firstnames":["Michael"],"suffixes":[]},{"propositions":[],"lastnames":["Stein"],"firstnames":["Benno"],"suffixes":[]}],"year":"2024","series":"Lecture Notes in Computer Science","volume":"14638","pages":"108–126","publisher":"Springer Nature Switzerland","address":"Cham","doi":"10.1007/978-3-031-63536-6_7","abstract":"The increasing usage of social networks has led to a growing number of discussions on the Internet that are a valuable source of argumentation that occurs in real time. Such conversations are often made up of a large number of participants and are characterized by a fast pace. Platforms like X/Twitter and Hacker News (HN) allow users to respond to other users' posts, leading to a tree-like structure. Previous work focused on training supervised models on datasets obtained from debate portals like Kialo where authors provide polarity labels (i.e., support/attack) together with their posts. Such classifiers may yield suboptimal predictions for the noisier posts from X or HN, so we propose unsupervised prompting strategies for large language models instead. Our experimental evaluation found this approach to be more effective for X conversations than a model fine-tuned on Kialo debates, but less effective for HN posts (which are more technical and less argumentative). Finally, we provide an open-source application for converting discussions on these platforms into argument graphs.","isbn":"978-3-031-63536-6","langid":"english","url":"https://www.wi2.uni-trier.de/shared/publications/Lenz2024PolArgUnsupervisedPolarity.pdf","bibtex":"@inproceedings{Lenz2024PolArgUnsupervisedPolarity,\n title = {{{PolArg}}: {{Unsupervised Polarity Prediction}} of~{{Arguments}} in~{{Real-Time Online Conversations}}},\n shorttitle = {{{PolArg}}},\n booktitle = {Robust {{Argumentation Machines}}},\n author = {Lenz, Mirko and Bergmann, Ralph},\n editor = {Cimiano, Philipp and Frank, Anette and Kohlhase, Michael and Stein, Benno},\n year = {2024},\n series = {Lecture {{Notes}} in {{Computer Science}}},\n volume = {14638},\n pages = {108--126},\n publisher = {Springer Nature Switzerland},\n address = {Cham},\n doi = {10.1007/978-3-031-63536-6_7},\n abstract = {The increasing usage of social networks has led to a growing number of discussions on the Internet that are a valuable source of argumentation that occurs in real time. Such conversations are often made up of a large number of participants and are characterized by a fast pace. Platforms like X/Twitter and Hacker News (HN) allow users to respond to other users' posts, leading to a tree-like structure. Previous work focused on training supervised models on datasets obtained from debate portals like Kialo where authors provide polarity labels (i.e., support/attack) together with their posts. Such classifiers may yield suboptimal predictions for the noisier posts from X or HN, so we propose unsupervised prompting strategies for large language models instead. Our experimental evaluation found this approach to be more effective for X conversations than a model fine-tuned on Kialo debates, but less effective for HN posts (which are more technical and less argumentative). Finally, we provide an open-source application for converting discussions on these platforms into argument graphs.},\n isbn = {978-3-031-63536-6},\n langid = {english},\n url = {https://www.wi2.uni-trier.de/shared/publications/Lenz2024PolArgUnsupervisedPolarity.pdf}\n}\n\n","author_short":["Lenz, M.","Bergmann, R."],"editor_short":["Cimiano, P.","Frank, A.","Kohlhase, M.","Stein, B."],"key":"Lenz2024PolArgUnsupervisedPolarity","id":"Lenz2024PolArgUnsupervisedPolarity","bibbaseid":"lenz-bergmann-polargunsupervisedpolaritypredictionofargumentsinrealtimeonlineconversations-2024","role":"author","urls":{"Paper":"https://www.wi2.uni-trier.de/shared/publications/Lenz2024PolArgUnsupervisedPolarity.pdf"},"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://web.wi2.uni-trier.de/publications/WI2Publikationen.bib","dataSources":["MSp3DzP4ToPojqkFy","CnwPa99ZchEF4SZgh"],"keywords":[],"search_terms":["polarg","unsupervised","polarity","prediction","arguments","real","time","online","conversations","lenz","bergmann"],"title":"PolArg: Unsupervised Polarity Prediction of Arguments in Real-Time Online Conversations","year":2024}