A Multi-cluster Boundary Learning Method for Out-of-Scope Intent Detection via MiniLM Embedding
Quick Answer
This study introduces a multi-cluster boundary learning method for out-of-scope (OOS) intent detection using MiniLM embedding, achieving state-of-the-art performance on CLINC150, StackOverflow, and Banking77 datasets.
Quick Take
This study introduces a multi-cluster boundary learning method for out-of-scope (OOS) intent detection using MiniLM embedding, achieving state-of-the-art performance on CLINC150, StackOverflow, and Banking77 datasets. The approach improves detection accuracy by treating OOS detection as a one-class classification task, addressing limitations of traditional multi-class methods and large parameter models.
Key Points
- Proposes a novel method for OOS intent detection using MiniLM embedding.
- Achieves state-of-the-art performance on multiple benchmark datasets.
- Addresses challenges in traditional multi-class classification approaches.
- Utilizes a one-class classification workflow for improved accuracy.
- Code is available for further research and implementation.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Intent detection is a critical task that bridges human intents and system actions in human-machine interaction systems. However, there still exist challenges for detecting out-of-scope (OOS) intents. (i) The traditional methods view the OOS intent detection as a multi-class classification, then the detection accuracy decreases as the class number of the known intents increases; (ii) LLM-embedding methods require large parameters, that makes them difficult to train and practically deploy. Thus, this work proposes a multi-cluster boundary learning method to detect OOS intents via MiniLM embedding (i.e., all-MiniLM-L6-v2) in an one-class classification workflow. The method learns the boundaries of multi-cluster embeddings generated by MiniLM from the training utterances, and then rejects the out-of-domain utterances as OOS intents. Experiments are conducted on public CLINC150, StackOverflow and Banking77 datasets. The results show that the method achieves the state-of-the-art OOS intent detection performance compared the other baselines. Ablation studies are also conducted and the results show that the used MiniLM can better adapt to the workflow and utterance embedding requirements. The code is available at supplementary materials.
| Comments: | To submit |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2607.07974 [cs.CL] |
| (or arXiv:2607.07974v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2607.07974 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mingyu Kang [view email]
[v1]
Wed, 8 Jul 2026 22:59:43 UTC (4,764 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →Quantifying Prior Dominance in Systems
The study introduces the Normalized Context Utilization (NCU) metric to evaluate Retrieval-Augmented Generation (RAG) systems, revealing that Small Language Models (SLMs) outperform larger models in factual extraction. The findings indicate that traditional scaling laws yield diminishing returns, with a commercial API frequently failing against adversarial evidence due to systemic confidence collapse.