Parsa Kamalipour
MSc student in Computer Science
Algorithms & Complexity Lab, Concordia University
Parsa studies community detection in social and complex networks: finding the groups hidden inside a graph, and judging whether they are real, with tools from algorithms and graph theory. He is a thesis-based master’s student at Concordia University in Montréal, supervised by Professor Hovhannes Harutyunyan.
Research
His thesis, Geodesic-Compact Community Detection and Quality Assessment in Social Networks (expected 2027), starts from a simple idea: the members of a good community are close to each other in shortest-path distance, compared with the rest of the network. This gives a way to find and judge communities without ground-truth labels. The work has so far produced Spider, a near-linear community detection algorithm that combines local and global steps (Computers, 2026), and a first-author study of the weighted-average Geodesic Distance Modularity (wGDM) with Metric Backbone sparsification as preprocessing for Leiden (IEEE SNAMS 2025, presented in Vienna). Read the full biography
Selected publications
- Spider Community Detection: Seeded Geodesic Expansion with Modularity-Guided Refinement and Greedy Merge MatchingComputers, Special Issue: Recent Advances in Social Networks and Social Media, 2026
- From Dense Graphs to Meaningful Communities: Assessing Community Quality Using Geodesic Distance Modularity on Metric Backbone-Sparsified NetworksThe 12th International Conference on Social Networks Analysis, Management and Security (IEEE SNAMS2025), 2025
Recent news
Graduate teaching assistant for four courses this fall: COEN 311, COEN 317, SOEN 363 and COMP 339.
Received the Dean’s Excellence Award for Excellence in Teaching by a Teaching Assistant from the Gina Cody School of Engineering and Computer Science.
“Spider Community Detection: Seeded Geodesic Expansion with Modularity-Guided Refinement and Greedy Merge Matching” is published in Computers 15(2), article 83. Read the paper
“From Dense Graphs to Meaningful Communities” presented at IEEE SNAMS 2025 in Vienna.
- Publications
- 4
- Courses as TA
- 15+
- Teaching hours
- 1650+
- Research positions
- 3
Last updated October 23, 2026