Snuba: Automating Weak Supervision to Label Training Data

Summary: Snuba automates weak supervision by generating task-specific labeling heuristics from a small labeled set to label a large unlabeled corpus. It grows coverage iteratively with a statistical termination guarantee, finishing under five minutes and beating handcrafted rules by 9.74 F1 and semi-supervised baselines by 14.35 F1. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: hf201c174eb06c565
Venue: VLDB
Year: 2019
Pagerank: 0.00011946047
Overall Rank: 1,120 | 92.48%
DOI: 10.14778/3291264.3291268

Incoming Non-self Citations Over Time

Authors

1. Paroma Varma (Stanford University)
2. Christopher Ré (Stanford University)

BibTeX Citation

@article{varma_vldb19,
        title = {{Snuba: Automating Weak Supervision to Label Training Data}},
        author = {Varma, Paroma and Ré, Christopher},
        journal = {PVLDB},
        series = {{VLDB} '19},
        volume = {12},
        number = {3},
        pages = {223--236},
        doi = {10.14778/3291264.3291268},
        url = {https://doi.org/10.14778/3291264.3291268},
        year = {2019}
}

Incoming Citations (Sorted by Pagerank)

Showing 26 of 26 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
501	Language Models Enable Simple Systems for Generating Structured Views of Heterogeneous Data Lakes	2024	VLDB	0.00017267905
3,627	The Role of Massively Multi-Task and Weak Supervision in Software 2.0	2019	CIDR	7.1514557e-05
3,955	Smile: A System to Support Machine Learning on EEG Data at Scale	2019	VLDB	6.9025583e-05
4,543	GOGGLES: Automatic Image Labeling with Affinity Coding	2020	SIGMOD	6.5470116e-05
5,076	Rotom: A Meta-Learned Data Augmentation Framework for Entity Matching, Data Cleaning, Text Classification, and Beyond	2021	SIGMOD	6.2860582e-05
5,100	Towards Benchmarking Feature Type Inference for AutoML Platforms	2021	SIGMOD	6.27349e-05
5,178	Explainable AI: Foundations, Applications, Opportunities for Data Management Research	2022	SIGMOD	6.2411759e-05
5,850	Adaptive Rule Discovery for Labeling Text Data	2021	SIGMOD	5.9719174e-05
5,970	VOCAL: Video Organization and Interactive Compositional AnaLytics	2022	CIDR	5.9316412e-05
6,662	UPLIFT: Parallelization Strategies for Feature Transformations in Machine Learning Workloads	2022	VLDB	5.7171651e-05
6,937	Inspector Gadget: A Data Programming-based Labeling System for Industrial Images	2021	VLDB	5.6397218e-05
7,473	Witan: Unsupervised Labelling Function Generation for Assisted Data Programming	2022	VLDB	5.5168918e-05
7,876	CHEF: A Cheap and Fast Pipeline for Iteratively Cleaning Label Uncertainties	2021	VLDB	5.4348457e-05
8,664	CrowdGame: A Game-Based Crowdsourcing System for Cost-Effective Data Labeling	2019	SIGMOD	5.2926663e-05
8,690	Nemo: Guiding and Contextualizing Weak Supervision for Interactive Data Programming	2022	VLDB	5.2905577e-05
8,756	Exploratory Training: When Annotators Learn About Data	2023	SIGMOD	5.2840289e-05
9,036	LANCET: Labeling Complex Data at Scale	2021	VLDB	5.2332952e-05
9,708	Ground Truth Inference for Weakly Supervised Entity Matching	2023	SIGMOD	5.1374628e-05
10,043	The Image Calculator: 10x Faster Image-AI Inference by Replacing JPEG with Self-designing Storage Format	2024	SIGMOD	5.0921006e-05
10,112	Data Augmentation for ML-driven Data Preparation and Integration	2021	VLDB	5.0789354e-05
10,211	CORAL: Collaborative Automatic Labeling System based on Large Language Models	2024	VLDB	5.0596605e-05
10,945	Morphing-based Compression for Data-centric ML Pipelines	2026	VLDB	4.9793485e-05
11,174	A Cost-Effective LLM-based Approach to Identify Wildlife Trafficking in Online Marketplaces	2025	SIGMOD	4.9793485e-05
11,218	WeShap: Weak Supervision Source Evaluation with Shapley Values	2025	VLDB	4.9793485e-05
11,720	Steered Training Data Generation for Learned Semantic Type Detection	2023	SIGMOD	4.9793485e-05
12,125	Leveraging Organizational Resources to Adapt Models to New Data Modalities	2020	VLDB	4.9793485e-05

Outgoing Citations (Sorted by Pagerank)

Showing 5 of 5 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
205	Snorkel: Rapid Training Data Creation with Weak Supervision	2018	VLDB	0.00025181304
427	Big Data Integration	2013	VLDB	0.00018465558
1,284	Fusing Data with Correlations	2014	SIGMOD	0.00011202115
3,061	Fonduer: Knowledge Base Construction from Richly Formatted Data	2018	SIGMOD	7.6927483e-05
3,791	SLiMFast: Guaranteed Results for Data Fusion and Source Reliability	2017	SIGMOD	7.0189865e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	3,627	The Role of Massively Multi-Task and Weak Supervision in Software 2.0	2019	CIDR
2	4,543	GOGGLES: Automatic Image Labeling with Affinity Coding	2020	SIGMOD
3	11,213	Two Birds with One Stone: Efficient Deep Learning over Mislabeled Data through Subset Selection	2025	SIGMOD
4	6,165	Demonstration of Panda: A Weakly Supervised Entity Matching System	2021	VLDB
5	8,690	Nemo: Guiding and Contextualizing Weak Supervision for Interactive Data Programming	2022	VLDB
6	5,181	Snorkel DryBell: A Case Study in Deploying Weak Supervision at Industrial Scale	2019	SIGMOD
7	5,850	Adaptive Rule Discovery for Labeling Text Data	2021	SIGMOD
8	9,708	Ground Truth Inference for Weakly Supervised Entity Matching	2023	SIGMOD
9	3,736	Snorkel: Fast Training Set Generation for Information Extraction	2017	SIGMOD
10	205	Snorkel: Rapid Training Data Creation with Weak Supervision	2018	VLDB