Akito Monden

Akito Monden is a prolific Japanese software-engineering scholar with 171 publications recorded in dblp during 1995-2025. His recent work centres on defect prediction, online learning, bandit-based tool selection, and empirical studies of software quality and human factors.

Research interests span software defect prediction, mining software repositories, effort estimation, clone detection, code generation, and the application of machine-learning techniques (notably bandit algorithms and ensemble methods) to practical software-engineering tasks. He also investigates requirements ambiguity, security-bug identification with large language models, and gaze-based human-computer interaction in programming education.

Across 2022-25 articles Monden explores online learning to cope with concept drift in defect datasets, multi-armed bandit algorithms for dynamic selection of clone detectors, fault-localisation techniques and code generators, and LLM-based security-bug detection. These themes reflect a sustained focus on data-driven, adaptive approaches that improve software quality assurance processes.