Creator and Annotator, Igbo-English Hate Speech Dataset
OtherTextClassification
Created and labeled a novel 268-example dataset of Igbo-English code-mixed text for hate speech detection as a final year research project. Annotated each entry with hate speech, non-hate, and nuanced language use, defining clear labeling guidelines for the dataset. Developed and trained a Keras-based neural classifier using these labeled examples and shipped a Kivy GUI for real-time model evaluation.
• Built dataset from scratch, including manual text collection and annotation.
• Labeled for complex hate, non-hate, and code-mixed tokens.
• Actively participated in model validation using labeled data.
• Designed annotation schema and labeling workflow documentation.
Created and labeled a novel 268-example dataset of Igbo-English code-mixed text for hate speech detection as a final year research project. Annotated each entry with hate speech, non-hate, and nuanced language use, defining clear labeling guidelines for the dataset. Developed and trained a Keras-based neural classifier using these labeled examples and shipped a Kivy GUI for real-time model evaluation. • Built dataset from scratch, including manual text collection and annotation. • Labeled for complex hate, non-hate, and code-mixed tokens. • Actively participated in model validation using labeled data. • Designed annotation schema and labeling workflow documentation.
2022 - 2022