Trained in physical, mathematical (BSc Double Math & Physics, University of Punjab 1989), and computational sciences (BSc Computer Science, University of Saskatchewan 1993; MSc Computer Science, University of Saskatchewan 1995; PhD Computer Science, University of Alberta 1999), Dr. Upal has become a polymath (similar to his role models Roger Schank, Herbert Simon, and Allen Newell) with key contributions to:
Machine Learning and Data Sciences (a system for predicting social media message popularity),
Agent-based Simulations (knowledge rich agent-based social simulation technology)
Rational Choice Theory & Social Identity Theory (a synthesis of RIT & SIT and development of the Social Identity Entrepreneurship model of leadership),
Cognitive Science of Religion (context-based model of counterintuitive concept learning & development of a Cognitive Science of New Religious Movements), and
Marketing (development of an ad design technique dubbed Aha ads).
A System to Predict the Number of Likes/Shares a Message is likely to Receive from a Target Audience
For his masters thesis, Dr. Upal worked under the supervision of probabilistic reasoning expert Dr. Eric Neufeld to develop a systematic methodology for comparing comparing clustering algorithms. His PhD thesis involved developing a novel machine learning algorithm to improve the performance of automated planning systems (such as a route planning algorithms). He continued to develop novel machine learning techniques for various applications such as RoboSoccer during his tenure as an Assistant Professor at Dalhousie University. Recently, Dr. Upal has developed a novel machine learning technique to automatically compute the popularity (likes/favorites and shares/retweets) of a social media message in a given target audience. Read Dr. Upal's machine learning articles.
Knowledge-rich Agent-based Simulation Technology
When confronted with real world national security & defense problems as a Senior Scientist at the DARPA contractor Information Extraction & Transport (IET) Inc, Dr. Upal found the existing agent-based simulation and cognitive modeling/AI techniques to be lacking and pioneered the development of knowledge rich agent-based simulation techniques. He has continued this work under the sponsorship of US DoD, Canadian DND, and European Union. Most recently he has designed a system to simulate the dynamics of social identity beliefs in situations of collective violence. Read Dr. Upal's articles on knowledge-rich agent-based simulation technology.
A Synthesis of Rational Choice Theory & Social Identity Theory
One of the biggest advantages of computational modeling is that one is forced to specify a model at a far more detailed level than verbal models. During his attempts to build agent-based models of social identity theory, Dr. Upal realized that there were a number of ill specified areas. He filled these gaps by drawing from elements of rational choice theory to develop a detailed computational (and falsifiable) model of social identity change. Read Dr. Upal's articles on synthesis of rational choice theory & social identity theory.
Social Identity Entrepreneurship Model of Leadership
Building on the work of social identity scholars of leadership Dr. Upal further developed the model of leaders as sellers of ideas of social identity change. Read Dr. Upal's articles on social identity entrepreneurship model of leadership.
Context-based Model to Explain the Spread of Maximally Counterintuitive Concepts (such as God)
Given his graduate training in machine learning, Dr. Upal was fascinated by the claim by cognitive scientists of religion (CSR) such as Pascal Boyer that many religious concepts around the world are minimally counterintuitive (MCI) and that this is because MCI concepts are more memorable than intuitive concepts. The CSR models did not explain why evolution has led to the development of a memory system that favored MCI concepts. Dr. Upal developed a rigorous theoretical frameworks based on concept learning research (esp Roger Schank 99, Murphy & Medin 85, and Gopnik and Meltzoff 97) dubbed the context-based model. The model not only explains the spread of MCI ideas but also explains how maximally counterintuitive concepts such as the Judeo-Christian & Islamic concept of God and new religious movement (NRM) doctrine. Read Dr. Upal's articles on the context-based model of memory for counterintuitive concepts.
Development of Aha Ad Design Technique
Based on his research on memory for counterintuitive concepts, Dr. Upal developed an ad-design technique (dubbed Aha ads) for designing effective advertising messages that get a target audience to engage in desired behavior. The technique uses the element of surprise to get a target audience to spend their cognitive resources to discover the advertiser's message, thereby making is memorable (and therefore actionable) for them. Read Dr. Upal's articles on Aha ad design technique.