Quality Assessment of Tabular Data using Large Language Models and Code Generation
This work addresses data quality assessment for tabular datasets, offering a novel method that reduces human intervention and computational costs, though it appears incremental as it builds on existing LLM and clustering techniques.
The paper tackles the problem of inefficient and costly rule-based validation for tabular data quality by introducing a three-stage framework that combines statistical detection with LLM-driven rule and code generation, achieving effectiveness confirmed through extensive evaluations on benchmark datasets.
Reliable data quality is crucial for downstream analysis of tabular datasets, yet rule-based validation often struggles with inefficiency, human intervention, and high computational costs. We present a three-stage framework that combines statistical inliner detection with LLM-driven rule and code generation. After filtering data samples through traditional clustering, we iteratively prompt LLMs to produce semantically valid quality rules and synthesize their executable validators through code-generating LLMs. To generate reliable quality rules, we aid LLMs with retrieval-augmented generation (RAG) by leveraging external knowledge sources and domain-specific few-shot examples. Robust guardrails ensure the accuracy and consistency of both rules and code snippets. Extensive evaluations on benchmark datasets confirm the effectiveness of our approach.