Throughout this project, we got a lot of perspective on how we can make real, tangible change through use of AI applications. Although software like Grammarly exists now, and has an accurate method of determining your tone as you write, we have yet to see a similarly widespread app that can intuitively show us other people’s tone.
Testing different values for the n-gram range as well as the max features parameter in the Tfidfvectorizer model.
Spend more time fitting our data to a specific machine learning model and try each one at a time to maximize its effectiveness. use a more current dataset, with more common Twitter language and humour for better accuracy in the current day.
More thoroughly address commonly mispelled words and adjust our means of processing them.
Our instructors told us that data scientists spend the majority of their time cleaning up their data, and we totally understand that now.
Making accessible communication possible in a way that does not alienate those in need of it is vitally important in achieving social equity. Everyone deserves to connect with others on an emotional level, and in an age of mainly digital correspondence, there is a void for those that need help interpreting feelings through text. AI is an incredibly valuable and popular tool throughout major industries; but it also proves to be a tool with great potential for social change. One tagged tweet at a time, we can help make the internet an accessible place for all; thanks to AI for All!!