Using New York Times Picks to Identify Constructive Comments

Varada Kolhatkar, Maite Taboada


Abstract
We examine the extent to which we are able to automatically identify constructive online comments. We build several classifiers using New York Times Picks as positive examples and non-constructive thread comments from the Yahoo News Annotated Comments Corpus as negative examples of constructive online comments. We evaluate these classifiers on a crowd-annotated corpus containing 1,121 comments. Our best classifier achieves a top F1 score of 0.84.
Anthology ID:
W17-4218
Volume:
Proceedings of the 2017 EMNLP Workshop: Natural Language Processing meets Journalism
Month:
September
Year:
2017
Address:
Copenhagen, Denmark
Venue:
WS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
100–105
Language:
URL:
https://aclanthology.org/W17-4218
DOI:
10.18653/v1/W17-4218
Bibkey:
Cite (ACL):
Varada Kolhatkar and Maite Taboada. 2017. Using New York Times Picks to Identify Constructive Comments. In Proceedings of the 2017 EMNLP Workshop: Natural Language Processing meets Journalism, pages 100–105, Copenhagen, Denmark. Association for Computational Linguistics.
Cite (Informal):
Using New York Times Picks to Identify Constructive Comments (Kolhatkar & Taboada, 2017)
Copy Citation:
PDF:
https://aclanthology.org/W17-4218.pdf