Afaan Oromo Hate Speech Detection and Classification on Social Media

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 6612–6619 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 6612 Afaan Oromo Hate Speech Detection and Classification on Social Media Teshome Mulugeta Ababu, Michael Melese Woldeyohannis School of Computing, Dire Dawa University Institute of Technology, Dire Dawa, Ethiopia School of Information Science, Addis Ababa University, Addis Ababa, Ethiopia teshome.mulugeta@ddu.edu.et, michael.melese@aau.edu.et Abstract Hate and offensive speech on social media is targeted to attack an individual or group of community based on protected characteristics such as gender, ethnicity, and religion. Hate and offensive speech on social media is a global problem that suffers the community especially, for an under-resourced language like Afaan Oromo language. One of the most widely spoken Cushitic language families is Afaan Oromo. Our objective is to develop and test a model used to detect and classify Afaan Oromo hate speech on social media. We developed numerous models that were used to detect and classify Afaan Oromo hate speech on social media by using different machine learning algorithms (classical, ensemble, and deep learning) with the combination of different feature extraction techniques such as BOW, TF-IDF, word2vec, and Keras Embedding layers. To perform the task, we required Afaan Oromo datasets, but the datasets were unavailable. By concentrating on four thematic areas of hate speech, such as gender, religion, race, and offensive speech, we were able to collect a total of 12,812 posts and comments from Facebook. BiLSTM with pre-trained word2vec feature extraction is an outperformed algorithm that achieves better accuracy of 0.84 and 0.88 for eight classes and two classes, respectively. Keywords: Afaan Oromo, Hate and Offensive Speech, Under-resourced, Machine Learning, Deep Learning 1. Introduction Today, Natural language processing (NLP) and Machine learning (ML) are over simplify different natural language applications such as speech recognition, sentiment analy- sis, hate speech detection, and other related NLP applica- tions. According to the 2020 World Bank statistics report1, Ethiopia is one of the second largest populated countries in Africa next to Nigeria, which has a more than 114 mil- lion population. Oromo ethnic group spoke Afaan Oromo as a mother tongue language that is written using Latin script called Qubee Afaan Oromoo (Degeneh Bijiga, 2015). Ethiopia is a multilingual and multi-ethnic country. In Ethiopia, Oromo is the largest ethnic group that contributes more than 37 million of the Ethiopian population (Eber- hard, David M., Gary F. Simons, and Charles D. Fennig (eds.), 2021). Afaan Oromo is spoken in Ethiopia special in Oromia re- gional states and neighboring African countries like Kenya, Tanzania, Sudan, Somalia (Degeneh Bijiga, 2015). Afaan Oromo is the fourth most widely spoken language in Africa next to Arabic, Hausa, and Swahili (Degeneh Bijiga, 2015). Social media is an attractive computer-mediated technol- ogy that is used to facilitate communication and sharing of data interactively and effectively for more than one person at a time (FDRE, 2020). Today, people can communicate easily and quickly with each other by using social media like Facebook, YouTube, …

Content truncated to 500 words