StarDrinks: An English and Korean Test Set for SLU Evaluation in a Drink Ordering Scenario
This dataset addresses the lack of realistic, multilingual benchmarks for task-oriented SLU in a specific domain, but it is an incremental contribution as it focuses on a single scenario.
The authors introduce StarDrinks, a bilingual (English and Korean) test set for evaluating spoken language understanding in a drink ordering scenario, featuring realistic speech phenomena and annotated slots. The dataset supports ASR, NLU, and SLU evaluation, providing a benchmark for model robustness.
LLMs and speech assistants are increasingly used for task-oriented interactions, yet their evaluation often relies on controlled scenarios that fail to capture the variability and complexity of real user requests. Drink ordering, for example, involves diverse named entities, drink types, sizes, customizations, and brand-specific terminology, as well as spontaneous speech phenomena such as hesitations and self-corrections. To address this gap, we introduce StarDrinks, a test set in English and Korean containing speech utterances features, transcriptions, and annotated slots. Our dataset supports speech-to-slots SLU, transcription-to-slots NLU, and speech-to-transcription ASR evaluation, providing a realistic benchmark for model robustness and generalization in a linguistically rich, real-world task.