hate Comments Detection
-
Efficient Tech: I chose DistilBERT for its efficiency, retaining BERT's language understanding while being smaller and faster.
-
Streamlined Development: Leveraging KerasNLP simplifies NLP application development and fine-tune the model to classify text comments into multiple categories, including "toxic," "severe_toxic," "obscene," "threat," "insult," and "identity_hate.".
-
Main Objective: My primary goal is creating a model to automatically classify toxic comments in text data, promoting a safer online environment.
-
Data Foundation: I use the Jigsaw Toxic Comment Classification Challenge dataset for model training and evaluation.
-
Efficiency Focus: The project prioritizes efficient toxic comment identification without compromising performance.
-
Impact-Driven: I recognize the importance of addressing online toxicity and aim to contribute to a more respectful online community.