Anurag Acharya

School of Computing and Information Sciences

Lecture Information:
  • October 7, 2020
  • 12:30 PM
  • Zoom:

Speaker Bio

Anurag Acharya is a PhD candidate in the School of Computing and Information Sciences (SCIS) at Florida International University (FIU), under the supervision of Dr. Mark Finlayson. He is part of the Cognac laboratory (Cognition, Narrative, and Culture). His research interest falls within the intersection of Artificial Intelligence (AI), Natural Language Processing (NLP), Computational Linguistics, and Cognitive Science. Anurag has a B.E. in Computer Engineering and B.A. in English and Political Science from Tribhuvan University, Kathmandu, Nepal. Anurag also worked as a PhD Intern in the National Security Directorate of the Pacific Northwest National Laboratory (PNNL) earlier this year.


As the field of Artificial Intelligence (AI) and Natural Language Processing (NLP) has evolved, there has now been a significant amount of focus on how to make them more universal, intelligent, and overall more human-like. But the majority of the work in the field seems to miss out on one key thing that is inherently part of us humans: culture. This is one glaring omission. Our cognition is shaped quite a bit by culture. Most of our perspective and language is shaped by culture as well. It is hard to imagine being able to construct a human-like system when one of the factors that shape human cognition is not incorporated into the process. As we move forward, it is vital that we embed cultural information into AI systems. I propose that this begin using two different forms of cultural knowledge: rituals and motifs. Rituals are a ‚Äúculturally defined set of behavior”. Motifs are distinct, recurring narrative elements found in folklore and, more generally, cultural materials.

For my dissertation, I propose the two major thrusts of work with the ultimate aim of integrating cultural knowledge into artificially intelligent systems. First, I propose to embed cultural knowledge into commonsense reasoning for Question Answering systems using Rituals. Secondly, I propose to build a system that can provide associations and other key information as output when presented with a motif as the input. Both of these will be done in two major steps – first a cognitive psychology study to collect relevant data from in-culture people, and then to use the data to build a computational system.