SenZi: A Sentiment Analysis Lexicon for the Latinised Arabic (Arabizi)

Tobaili, Taha; Fernandez, Miriam; Alani, Harith; Sharafeddine, Sanaa; Hajj, Hazem and Glavas, Goran (2019). SenZi: A Sentiment Analysis Lexicon for the Latinised Arabic (Arabizi). In: International Conference Recent Advances In Natural Language Processing 2019 Natural Language Processing in a Deep Learning World: Proceedings (Angelova, Galia; Mitkov, Ruslan; Nikolova, Ivelina and Temnikova, Irina eds.) pp. 1204–1212.

Abstract

Arabizi is an informal written form of dialectal Arabic transcribed in Latin alphanumeric characters. It has a proven popularity on chat platforms and social media, yet it suffers from a severe lack of natural language processing (NLP) resources. As such, texts written in Arabizi are often disregarded in sentiment analysis tasks for Arabic. In this paper we describe the creation of a sentiment lexicon for Arabizi that was enriched with word embeddings. The result is a new Arabizi lexicon consisting of 11.3K positive and 13.3K negative words. We evaluated this lexicon by classifying the sentiment of Arabizi tweets achieving an F1-score of 0.72. We provide a detailed error analysis to present the challenges that impact the sentiment analysis of Arabizi.

Viewing alternatives

Item Actions

Export

You can export this page using these formats