Speaker: Segun Eroyehun

Title: Applications of Deep learning to social media mining

Abstract: The advent of social media has presented us with both opportunities and challenges as a society and for NLP. The user-generated content from social media platforms presents many challenges ranging from noise to limited context. In this setting, deep learning models do not perform well. I will outline some of the current issues and possible approaches to tackle them on the tasks of content moderation and pharmacovigilance.

Speaker: Jason Angel

Title: The use of lexical resources for natural language processing

Abstract: Most of the current NLP models rely on distributional representation of words. We show an alternative approach for NLP tasks using lexical resources and deep learning. I will present a generic pipeline that demonstrates the integration of lexical resources for NLP tasks, with results on patent classification and bridging anaphora resolution.