Framework for Affective News Analysis of Arabic News: 2014 Gaza Attacks Case Study

Mahmoud Al-Ayyoub (Jordan University of Science and Technology, Jordan)

Huda Al-Sarhan (Jordan University of Science and Technology, Jordan)

Majd Al-So'ud (Jordan University of Science and Technology, Jordan)

Mohammad Al-Smadi (Jordan University of Science and Technology, Jordan)

Yaser Jararweh (Jordan University of Science and Technology, Jordan)

Abstract: This paper aims at fostering the domain of Arabic affective news analysis through providing: (a) a benchmark annotated Arabic dataset of news for affective news analysis, (b) an aspect-based sentiment analysis (ABSA) approach for evaluating the sentimental affect of Arabic news posts on the reader, and (c) a baseline approach with a common evaluation framework to compare future research results with the baseline ones.

Keywords: Arabic dataset, affective news, aspect based sentiment analysis, emotional affect, natural language processing

Categories: H.3, H.3.5, H.4.3, I.2.7, J.4