4 approaches applied to analyze various types of financial text

All code needs as input: text data and relevant earnings information

Notes on the approaches:

Skip gram model - is based on the work in the paper https://www.sciencedirect.com/science/article/pii/S0167923616300823

FastAI models - notebooks to utilize the models built in https://github.com/fastai/fastai/tree/master/courses/dl2/imdb_scripts

Sentiment Analysis - is based on modified versions on VADER located at https://github.com/cjhutto/vaderSentiment

CNN Models - notebook with a keras based model of the work in the paper https://arxiv.org/abs/1408.5882