DATE: Thu, Feb 16, 2017
TIME: 1 pm
PLACE: SITE 5084
TITLE: Using Cognitive Computing to Get Insights on Personality Traits from Twitter Messages
PRESENTER: Romualdo Alves Pereira Junior
University of Brasilia and University of Ottawa
ABSTRACT:

In this work, we use text mining algorithms on Twitter messages made available by PAN2015. In the rst step, we applied IBM Watson algorithms to obtain the results of Big Five personality analysis, here namely OCEAN: Openness, Conscientiousness, Extraversion, Agreeableness and Neuroticism. Then, we applied a Deep Learning algorithm on the resulted IBM Watson scores, in order to minimize the Root Mean Square Error. In this way, we achieved better results than using the IBM Watson algorithms alone. The dataset contains messages in English from 152 distinct authors. Additionally, we performed an analysis of personality traits to detail each component of OCEAN. Our results can be visualized in a Business Intelligence system.