ILab-Edinburgh at SemEval-2016 task 7: A hybrid approach for determining sentiment intensity of Arabic Twitter phrases

Abstract

This paper describes the iLab-Edinburgh Sentiment Analysis system, winner of the Arabic Twitter Task 7 in SemEval-2016. The system employs a hybrid approach of supervised learning and rule-based methods to predict a sentiment intensity (SI) score for a given Arabic Twitter phrase. First, the supervised method uses an ensemble of trained linear regression models to produce an initial SI score for each given text instance. Second, the resulting SI score is adjusted using a set of rules that exploit a number of publicly available sentiment lexica. The system demonstrates strong results of 0.536 Kendall score, ranking top in this task.