# Swahili-misogyny-detector
In a world where the internet shaped our interactions, detecting misogyny and sexism is important to create respectful digital communities. This project takes advantage of advanced language processing to translate Swahili to English and identify misogyny as well as encouraging users to intervene proactively. The results displayed are saved in a database (both the text input, the label and translation) for the creation of a Swahili dataset to be used in the continued development of a model customized for Swahili language. By use of different methodologies, such as pretrained models and retraining them with a new dataset, a system architecture has been developed that integrates user interaction, processing and data management. The user-friendly interface ensures easy engagement and a simple look has been used for easy navigation throughout the web app. This tool evolves to continually and effectively tackle misogyny therefore a step closer in promoting inclusive and safe online spaces.