Lars Bungum

Lars Bungum is a Researcher in the Department of Computer Science at NTNU Trondheim. His primary focus lies in Machine Translation (MT), Natural Language Processing (NLP), and Domain Adaptation techniques. He completed his doctoral dissertation in 2021 titled "Unsupervised Clustering of Structured and Unstructured Text Collections".

Key research interests include unsupervised clustering methods, cross-domain adaptation for MT systems, and computational linguistics challenges such as negation scope detection and semantic similarity. His work spans both theoretical advancements (e.g., agent-based modeling of language evolution) and practical applications like Twitter sentiment analysis and Norwegian noun compound translation.

Bungum has collaborated extensively with Prof. Björn Gambäck and other researchers on projects involving corpus modeling (e.g., PRESEMT Deliverable D3.3.2), statistical MT, and neural language models. His contributions often bridge foundational NLP research with real-world translation challenges, emphasizing scalable solutions for large text corpora.

Recent outreach activities include presenting at Nodalida 2023 on multilingual grammatical error detection and participating in panels on machine translation ethics. His research consistently addresses gaps in domain adaptation, corpus utilization, and post-editing strategies for literary texts.