Can Foundation Models Wrangle Your Data?

By Avanika Narayan, Ines Chami, Laurel J. Orr, Christopher R'e. Proceedings of the VLDB Endowment, 2022

344 citations recorded by the source corpus.

Foundation Models (FMs) are models trained on large corpora of data that, at very large scale, can generalize to new tasks without any task-specific finetuning. As these models continue to grow in size, innovations continue to push the boundaries of what these models can do on language and image tasks. This paper aims to understand an underexplored area of FMs: classical data tasks like cleaning and integration. As a proof-of-concept, we cast five data cleaning and integration tasks as prompting tasks and evaluate the performance of FMs on these tasks. We find that large FMs generalize and achieve SoTA performance on data cleaning and integration tasks, even though they are not trained for these data tasks. We identify specific research challenges and opportunities that these models present, including challenges with private and domain specific data, and opportunities to make data management systems more accessible to non-experts. We make our code and experiments publicly available at: https://github.com/HazyResearch/fm_data_tasks.

View Paper: https://www.semanticscholar.org/paper/58aacb967cc7fc25cfc9d51b7ad3e57ac00d119b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Christopher R'e
2. Lianmin Zheng
3. Ce Zhang
4. Ines Chami
5. Ying Sheng
6. Percy Liang
7. Beidi Chen
8. Daniel Y. Fu

Citations in the corpus

• TabLLM: Few-shot Classification of Tabular Data with Large Language Models (2022, unknown): “Narayan et al. (2022) recently assessed in-context learning with the autoregressive language model GPT-3 for tabular data cleaning tasks.”
• High-throughput Generative Inference of Large Language Models with a Single GPU (2023, methodology): “, 2018), data wrangling (Narayan et al., 2022), and form processing (Chen et al.”
• Ask Me Anything: A simple strategy for prompting language models (2022, background): “We hope AMA and future work help address painpoints of using LLMs [Arora and Ré, 2022, Narayan et al., 2022] by improving the ability to proceed with less-than-perfect prompts and enabling the use of small, private, and open-source LLMs.”
• ArchRAG: Attributed Community-based Hierarchical Retrieval-Augmented Generation (2025, unknown): “RAG has been proven to excel in many tasks, including open-ended question answering [28, 57], programming context [9, 10], SQL rewrite [38, 59], and data cleaning [46, 47, 51].”