Sentiment Analysis and the Enron Email Data Set
John Harris and Brian Levey
Analytical initiatives are no longer limited to structured or semi-structured data. The next iteration of web technologies and applications remains fairly undefined, but will likely involve a high volume of unstructured textual data. Social media, email and other forms of communication are driving opportunities by way of semantic analysis.
We will present specific analysis of the Enron email data set and use specific applications within this domain to cover relevant topics, considerations, and avenues for growth in the field of sentiment analysis. We will both cover analysis of this data set from a slightly different perspective and cover our findings in conjunction in order to generate discussion. Furthermore, John will present details of an active product development and startup opportunity that he pursuing and will discuss elements of this initiative including the startup process itself, challenges encountered, technical methods and business concepts learned along the way.
Boxelder Consulting was founded by John this year and was spun out of his product development company, ScratchPad Ventures. John’s focus at Boxelder is the collection, processing, analysis and distribution of data products and services that enhance decision-making and understanding for clients. John’s specific and most recent passion lies in the processing, analysis and delivery of unstructured textual data products in order to generate understanding of client business operations or their market in general.
Dr. Brian Levey works as a Senior Research Scientist at Strategic Analaysis Enterprises where his responsibilities include model generation, model transparency, research design, test and evaluation, and statistical analysis of data. He also acts as liaison with William and Mary, organizing and managing our research assistants. Dr. Levey received his Ph.D. in Political Science with a focus on social science methodology from the University of Georgia in 2013. Dr. Levey brings a strong political science and statistics background to the team and has been unofficially named the Chief of Randomness. His areas of interest include research design, measurement models, counterfactual analysis, and Bayesian statistics.