Though 2021 wasn’t an easy year for the School of Computing – with the transition to remote learning and working – our students and faculty continued to demonstrate their excellence. The Big Data Analytics and Management (BAM) Lab had three papers that won awards in the past year, a testament to the hard work and perseverance of the researchers.
The first paper was from Tariq Abughofa and Ahmed Harby, who won the 2021 IEEE Big Data Service Conference’s Best Paper Award for their paper “Incremental Community Detection in Distributed Dynamic Graph”. This publication outlines Tariq’s research about a new algorithm for efficiently updating a graph data structure with new incoming streaming data to detect strongly connected community of nodes. The technique is applicable to any connected or linked data such as social media networks, Internet of Things or connected devices, and data storage systems. The research was funded by GnowIt and SOSCIP (Smart Computing for Innovation).
The second paper was written by Jason Lam and Yuhao Chen, who won the IEEE International Conference on Data Mining and Learning in the Legal Domain (MLLD) 2021 best paper award for their work on “Detection of Similar Legal Cases on Personal Injury”. Their paper uses natural language processing and deep learning models to analyze unstructured text data from legal case descriptions, allowing for easier search for similar legal cases, an important aspect of Canada’s case law system. The paper specifically looks at cases involving personal injury and concludes that semantic similarity detection is not enough for legal case data, and artificial intelligence and machine learning are critical in finding case similarity using legal criteria. The research was funded by NSERC, NFRF and was performed in collaboration with Queen’s Law.
The third paper was written by Isaac Hogan, Donghao Qiao, Ruikang Luo, Mojtaba Moattari and Austin Carthy, who won the IEEE International Conference on Cognitive Machine Intelligence (CogMI 2021) best student paper award for their work on “FireWarn: Fire Hazards Detection Using Deep Learning Models”. Their paper posits that video cameras can be effectively used in surveillance systems to detect hazardous situations with the help of deep-learning techniques. The model created by the researchers achieved a 97% testing accuracy in labelling regions of interest in videos that show images of smoke and fire. This work was performed by the group of graduate and undergraduate researchers in collaboration with DRDC (Defense Research and Development Canada) and Royal Military College. These are just a few examples of the amazing work that happens every day here at the School of Computing. Congratulations to the students who won!