Text-to-SQL Task-oriented Dialogue Ontology Construction
This addresses the need for explainable and controllable LLMs in task-oriented dialogue systems, though it is incremental as it builds on existing SQL and dialogue theory.
The paper tackles the problem of manually building ontologies for task-oriented dialogue systems by introducing TeQoDO, a method that uses an LLM to autonomously construct ontologies from scratch without supervision, outperforming transfer learning approaches and achieving competitive results on downstream tasks.
Large language models (LLMs) are widely used as general-purpose knowledge sources, but they rely on parametric knowledge, limiting explainability and trustworthiness. In task-oriented dialogue (TOD) systems, this separation is explicit, using an external database structured by an explicit ontology to ensure explainability and controllability. However, building such ontologies requires manual labels or supervised training. We introduce TeQoDO: a Text-to-SQL task-oriented Dialogue Ontology construction method. Here, an LLM autonomously builds a TOD ontology from scratch without supervision using its inherent SQL programming capabilities combined with dialogue theory provided in the prompt. We show that TeQoDO outperforms transfer learning approaches, and its constructed ontology is competitive on a downstream dialogue state tracking task. Ablation studies demonstrate the key role of dialogue theory. TeQoDO also scales to allow construction of much larger ontologies, which we investigate on a Wikipedia and ArXiv dataset. We view this as a step towards broader application of ontologies to increase LLM explainability.