"VSLIM: A Vietnamese Explicit Slot-Intent Mapping for Joint Multi-Intent Detection and Slot Filling," by Phong Chung, Kha Le-Minh, Dr. Xuan-Bach Le and Assoc. Prof. Tho Quan, was published at ACIIDS 2026 (Asian Conference on Intelligent Information and Database Systems) in Kaohsiung, Taiwan, April 2026. Multi-intent detection and slot filling are foundational to natural language understanding in task-oriented dialogue. Early approaches treated them as separate problems, which severs the direct connection between an intent and the slots that belong to it — a limitation that gets worse when a single utterance carries several intents at once. In Vietnamese, work on this remains scarce, largely because of the language's low-resource status. VSLIM builds on the SLIM framework and adds a biaffine classifier that captures the intent–slot relationship more directly, letting the model represent dependencies across sequence labels in multi-intent settings. Experiments on the Vietnamese PhoATIS dataset and on VPED, a new corpus introduced by the authors, show VSLIM outperforming strong NLU baselines. Paper: https://lexuanbach.github.io/publication/ACIIDS2026b.pdf Poster: https://lexuanbach.github.io/slides/ACIIDS2026b_poster.pdf Code and data: https://github.com/dongphong543/VSLIM DOI: https://doi.org/10.1007/978-981-92-0263-8_3