Deep Dive was born at TechTogether Boston and is a social media background check for HR recruiters that uses natural language processing to identify hate speech and offensive language in a potential applicant's twitter.
The original dataset used for training can be found here: https://data.world/thomasrdavidson/hate-speech-and-offensive-language
Model implementation was inspired by Yoon Kim's paper: Convolutional Neural Networks for Sentence Classification (https://www.aclweb.org/anthology/D14-1181) and the GloVe pretrained Twitter word embeddings were used.
