Adverse Drug Reaction Classification using NLP tools and Deep Learning Methods
Our primary intent is to explore the extent to which ADR assertive text segments can be classified from text based data sources, particularly social media sources and structured datasets. The following list summarizes our intents: (i) Investigation of different Neural Network (NN) architectures for ADR classification. (ii) Exploration of NLP techniques to extract informative and portable features from text coming from distinct sources. (iii) Investigate the performance of supervised classification approaches on data from social media to data from other more structured and unstructured sources. (iv) Optimize machine learning and deep learning algorithms to improve performance over existing approaches.