hate.criME v1.1

Moving forward, I hope to incorporate natural language processing or predictive text with my analysis of hate crime news.

Natural Language Processing ToolKits may struggle with sensitive topics; for example, Google’s Cloud Natural Language API assigns negative sentiment to some ethnicities, religions, sexual orientations, and gender identities. Performing a sentiment analysis of hate crime articles could provide a powerful illustration of this shortcoming.

Predictive text could be used to make seemingly benign words relate to hurtful hate crime news headlines. The product could come in the form of a parody of a BuzzFeed quiz, in which the user inputs good-natured affinities and personal characteristics which then output some representation of hate crime news.