Financial behavior has long been a topic of interest in the fields of economics, psychology, and neuroscience. However, most studies on financial behavior are either studying macro trends or are in the forms of small-scale studies. This is due to the sensitive nature of financial records, which leads to a lack of large-scale dataset on financial activities. The landscape has been slowly changing. Since the emergence of ecommerce websites and the fast adoption of mobile wallets, more and more payments are conducted online. With the advent of cryptocurrencies, financial activities are even more deeply rooted in the online world.
These advancements provide rich opportunities for data-driven researches into financial behaviors. My research focuses on using data analysis and modeling techniques to extract diverse patterns in online financial behavior and accurately predict future behavior. I collect financial decisions and discussions from multiple large online platforms. The first study focuses on penny auctions in DealDash, where I identify the current winners and their behavioral patterns. In addition, I will demonstrate how to learn winning strategies by building a highly accurate auction simulator using recurrent neural networks. In the second part, I will discuss how we model the topics of interest among cryptocurrency users on Reddit and how such interests are indicative of their future behavior.