SemEval2019 Task 4 - Hyperpartisan News Detection

• Team's name: Fearnando-Pessa.
• Task overview: https://pan.webis.de/semeval19/semeval19-web/index.html
• Leaderboard: https://pan.webis.de/semeval19/semeval19-web/leaderboard.html

Task

Given a news article text, decide whether it follows a hyperpartisan argumentation, i.e., whether it exhibits blind, prejudiced, or unreasoning allegiance to one party, faction, cause, or person.

System's Performance

Model	Accuracy	Precision	Recall	F1
Random Forest (Official)	71.7	80.6	57.0	66.8
Gradient Boosted Trees (Best)	72.9	78.1	63.7	70.2
Baseline	46.2	46.0	44.3	45.1

Team Members

• André Cruz
• Gil Rocha
• Rui Sousa-Silva
• Henrique Lopes Cardoso

Citation

@inproceedings{cruz-etal-2019-team,
    title = "Team Fernando-Pessa at {S}em{E}val-2019 Task 4: Back to Basics in Hyperpartisan News Detection",
    author = "Cruz, Andr{\'e}  and
      Rocha, Gil  and
      Sousa-Silva, Rui  and
      Lopes Cardoso, Henrique",
    booktitle = "Proceedings of the 13th International Workshop on Semantic Evaluation",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/S19-2173",
    doi = "10.18653/v1/S19-2173",
    pages = "999--1003",
}