NIGERIAN BASED HATE DATASET
Published: 29 January 2024| Version 1 | DOI: 10.17632/r5jsynhxsx.1
Contributors:
Bassey Adim, , , Description
The dataset was built from tweeter and contains tweets based on common Nigerian hate words and stereotypes. The dataset features 20,176 tweets that have been classified into binary class: Hate Speech with polarity of 1, containing 4,801 tweets and Non Hate Speech with polarity 0, having 15,375 tweets. The dataset is presented in text csv table format, arranged in the following columns: 1. ʽId’ which gives the serial number of the tweets; 2. ʽTweets’. This Contain the contents of the tweets; and 3. 'Polarity' contains the label of tweets. Hate Speech have a polarity of 1 and Non-Hate Speech have a polarity of 0.
Files
Institutions
University of Lagos, Cross River University of Technology
Categories
Natural Language Processing, Machine Learning, Sentiment Analysis