Exploring the frontiers of artificial intelligence, knowledge systems, and human-centered computing. We build intelligent systems that understand, reason, and interact with the world.
github.com/CogNet-Lab
Exploring the frontiers of artificial intelligence, knowledge systems, and human-centered computing. We build intelligent systems that understand, reason, and interact with the world.
github.com/CogNet-LabResearch Focus
Onto-KGR
Advancing knowledge graph representation learning through ontology-guided construction and embedding methods. This line integrates multidisciplinary ontology building with state-of-the-art negative sampling strategies, textual knowledge distillation, and tensor decomposition techniques to produce structurally richer and semantically grounded KG embeddings across domains such as healthcare, governance, and enterprise systems.
AI4SE-Lab
Investigating how AI transforms software engineering practices — from autonomous quality assurance and intelligent defect prediction to real-time anomaly detection in API-driven and cloud-native systems. This project combines LLM-powered agents for automated test generation, self-healing test suites, and adaptive CI/CD pipelines with behavioral modeling techniques that autonomously detect and mitigate cyber threats in modern distributed architectures.
AdaptHCI
Investigating how intelligent systems can dynamically adapt their interface elements — tooltips, guidance overlays, and interaction patterns — based on user cognitive profiles, task context, and real-time behavior signals. Building on prior work in metacognition-aware design, this project develops adaptive UI frameworks that personalize information delivery to reduce cognitive load and improve accessibility across diverse user populations.
BlockTrust-KP
Converging blockchain-based trust infrastructure with semantic knowledge systems to create verifiable, tamper-proof provenance chains for AI-generated knowledge and digital assets. This direction combines decentralized governance frameworks with ontology-driven trust scoring, enabling transparent audit trails for DRM, public records, and AI reasoning outputs in enterprise and government contexts.
GreenSE-Carbon
Measuring and comparing the environmental impact of software systems across architectural styles and technology stacks. This project evaluates energy consumption and carbon emissions across monolithic, microservices, and event-driven architectures in containerized and bare-metal deployments using hardware-level profiling (HWiNFO64) and standardized load testing (k6). A complementary benchmarking tool compares the carbon footprint of popular web frameworks across multiple programming language ecosystems under controlled conditions.
DataBench-Lab
Developing research-grade, extensible benchmarking frameworks for evaluating database and datastore performance across multiple dimensions. Current work spans vector database evaluation for embedding-powered AI applications and comparative analysis of NoSQL systems (Cassandra, MongoDB, Redis) in distributed deployment modes — enabling fair, reproducible comparisons for both academic research and production system selection.
Scholarly Output
| Year | Title | Authors | Venue |
|---|---|---|---|
| 2026 | Multi-threaded Recast-Based A* Pathfinding for Scalable Navigation in Dynamic Game Environments | T Madushanka, S Madushanka | arXiv preprint arXiv:2602.04130 |
| 2024 | Negative Sampling in Knowledge Graph Representation Learning: A Review | T Madushanka, R Ichise | arXiv preprint |
| 2024 | SecureRights: A Blockchain-Powered Trusted DRM Framework for Robust Protection and Asserting Digital Rights | T Madushanka, DS Kumara, AA Rathnaweera | arXiv preprint arXiv:2403.06094 |
| 2024 | Text-KG Distill: Textual Knowledge Distillation for Enhanced KG Embeddings | T Madushanka, R Ichise | 2024 International Conference on AI x Data and Knowledge Engineering (AIxDKE) |
| 2024 | Blockchain for Vehicle Registration, Transferring and Management Process in Sri Lanka | C Malintha, D Diyasena, T Madushanka | SSRN |
| 2023 | TuckerDNCaching: High-Quality Negative Sampling with Tucker Decomposition | T Madushanka, R Ichise | Journal of Intelligent Information Systems, 61(3), 739–763 |
| 2022 | MDNCaching: A Strategy to Generate Quality Negatives for Knowledge Graph Embedding | T Madushanka, R Ichise | International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems |
| 2022 | A Novel Anomaly Detection Approach to Secure APIs from Cyberattacks | A Ifthikar, N Thennakoon, S Malalgoda, HK Moraliyage, T Jayawickrama, T Madushanka, et al. | La Trobe University |
| 2020 | Impact of Metacognition and Age Group on Contemporary Video Game Interface and Gameplay Design | NN Harischandra, LA Jayakody, T Madusanka | 2020 International Research Conference on Smart Computing and Systems Engineering (SCSE) |
| 2020 | Building Automation System to Optimize Energy Utilization Acquiring the Best Performance of Appliances in a Distributed Network | I Ranaweera, T Madushanka, C Ranaweera | Proceedings of International Conference on Advances in Computing and Technology |
| 2016 | Performance Comparison of NoSQL Databases in Pseudo Distributed Mode: Cassandra, MongoDB & Redis | C Kumarasinghe, D Liyanage, T Madushanka, L Mendis | Conference Proceedings |
| 2015 | Customer Churn Prediction: A Cognitive Approach | D Senanayake, L Muthugama, L Mendis, T Madushanka | International Journal of Computer, Electrical, Automation, Control and Information Engineering |
We're always looking for motivated researchers, students, and collaborators who share our passion for intelligent systems, knowledge engineering, and human-centered AI. If you're interested in pursuing research in any of our focus areas, we'd love to hear from you.