Presentation of fellowship projects

Deep learning-based fake review detection model focusing on emotional expressions and explicit aspects

Ajay KUMAR  (emlyon business school)

April 22, 2020 – 12:00 pm- 1:00 pm

Abstract: In this research, we are trying to investigate the behavior of fake accounts on Yelp data set to generate fake negative/positive reviews for competitors or itself and identify the fake reviews based on several « review-centric » and « reviewer-centric » features, which contribute the most fake reviews in entire data set. It has been observed that the existing machine learning methods fail to classify the fake reviews and provide the low classification results. Therefore, we are trying to propose a deep learning-based fake review detection framework (FRDF) that detects fake reviews after extracting some unique and novel features from the data set and employ their collective behavior in a unified manner.

Biography: Ajay Kumar is Assistant Professor in the AIM Research Center. His research and teaching interests are in data and text mining, decision support systems, knowledge management, business intelligence and entreprise modeling. He worked as post doctoral scientist in business analytics & information systems domain at Massachusetts Institute of Technology and as postdoctoral fellow at Harvard Business School. He received his PhD from IIT Delhi where he worked to develop the big data driven predictive models in TLC and manufacturing sectors.

Bias in Artificial Intelligence

Hanan SALAM  (emlyon business school)

April 24, 2020 – 12:00 pm- 1:00 pm

Abstract: Applying AI to business application gives us the capacity to automate repetitive tasks as well as to increase business revenues by adapting to users’ needs, profiles and behaviors. However, machine learning data, algorithms, and other design choices that shape AI systems may reflect and amplify existing cultural bias and prejudices. Consequently, when we develop data-driven AI products and feed them data, we should be careful that we are neither simply confirming existing biases nor introducing new ones. This research will help advancing the state-of-art from heuristic repairing, to proactive and theoretically supported prevention tackling the bias problematic from both an algorithmic and a design perspectives

Biography: Hanan Salam holds a PhD in Artificial Intelligence. She is Co-founder of Women in AI, a non-profit Do-Tank whose mission is to close the gender gap in the domain of Artificial Intelligence through education, research and events. Her scientific interests include Artificial Intelligence, human-computer interaction, social robotics, computer vision, machine learning and affective computing. She is an advocate of technology for common good and an activist for women empowerment.