Observatory: Characterizing Embeddings of Relational Tables

Summary: Observatory introduces eight measurable properties grounded in relational invariants and data distributions to systematically characterize table embeddings. Evaluating nine models, it exposes sensitivity to column order, weak functional-dependency preservation, and lower sample fidelity in specialized models. (summarized by gpt-5.6-luna on Jul 24 2026)

Paper ID: h519478370a00cd73
Venue: VLDB
Year: 2024
Pagerank: 6.1809951e-05
Overall Rank: 5,322 | 64.24%
DOI: 10.14778/3636218.3636237
PDF: Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

Authors

1. Tianji Cong (University of Michigan)
2. Madelon Hulsebos (University of Amsterdam)
3. Zhenjie Sun (University of Michigan)
4. Paul Groth (University of Amsterdam)
5. H. V. Jagadish (University of Michigan)

BibTeX Citation

@article{cong_vldb24,
        title = {{Observatory: Characterizing Embeddings of Relational Tables}},
        author = {Cong, Tianji and Hulsebos, Madelon and Sun, Zhenjie and Groth, Paul and Jagadish, H. V.},
        journal = {PVLDB},
        series = {{VLDB} '24},
        volume = {17},
        number = {4},
        pages = {849--862},
        doi = {10.14778/3636218.3636237},
        url = {https://doi.org/10.14778/3636218.3636237},
        year = {2024}
}

Incoming Citations (Sorted by Pagerank)

Showing 5 of 5 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
658	DocETL: Agentic Query Rewriting and Evaluation for Complex Document Processing	2025	VLDB	0.00015058738
10,169	Cents: A Flexible and Cost-Effective Framework for LLM-Based Table Understanding	2025	VLDB	5.0658661e-05
10,826	FedAugment: Table Augmentation Search over Decentralized Data Repositories	2026	VLDB	4.9769913e-05
11,264	Birdie: Natural Language-Driven Table Discovery Using Differentiable Search Index	2025	VLDB	4.9769913e-05
11,433	Panel on Neural Relational Data: Tabular Foundation Models, LLMs... or both?	2025	VLDB	4.9769913e-05

Outgoing Citations (Sorted by Pagerank)

Showing 17 of 17 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
10	Extending the Data Base Relational Model to Capture More Meaning	1979	SIGMOD	0.00075910134
134	Deep Entity Matching with Pre-Trained Language Models	2021	VLDB	0.00030042569
329	Can Foundation Models Wrangle Your Data?	2023	VLDB	0.00020867521
367	InfoGather: Entity Augmentation and Attribute Discovery By Holistic Matching with Web Tables	2012	SIGMOD	0.00019876273
377	TURL: Table Understanding through Representation Learning	2021	VLDB	0.00019564011
603	A Hybrid Approach to Functional Dependency Discovery	2016	SIGMOD	0.00015692059
694	JOSIE: Overlap Set Similarity Search for Finding Joinable Tables in Data Lakes	2019	SIGMOD	0.00014721161
765	Table Union Search on Open Data	2018	VLDB	0.00014119196
1,319	LSH Ensemble: Internet-Scale Domain Search	2016	VLDB	0.00011042864
1,391	Creating Embeddings of Heterogeneous Relational Datasets for Data Integration Tasks	2020	SIGMOD	0.00010812249
1,533	On Multi-Column Foreign Key Discovery	2010	VLDB	0.00010328081
1,778	Annotating Columns with Pre-trained Language Models	2022	SIGMOD	9.6578031e-05
1,929	CHORUS: Foundation Models for Unified Data Discovery and Exploration	2024	VLDB	9.3551286e-05
1,993	RPT: Relational Pre-trained Transformer Is Almost All You Need towards Democratizing Data Preparation	2021	VLDB	9.2348951e-05
2,093	Sato: Contextual Semantic Type Detection in Tables	2020	VLDB	9.0593586e-05
4,631	Transformers for Tabular Data Representation: A Tutorial on Models and Applications	2022	VLDB	6.4926848e-05
7,164	WarpGate: A Semantic Join Discovery System for Cloud Data Warehouses	2023	CIDR	5.5934028e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	1,977	Table-GPT: Table Fine-tuned GPT for Diverse Table Tasks	2024	SIGMOD
2	6,022	Tables As a Paradigm for Querying and Restructuring	1996	PODS
3	9,254	Making Table Understanding Work in Practice	2022	CIDR
4	2,520	GitTables: A Large-Scale Corpus of Relational Tables	2023	SIGMOD
5	10,749	Database Views as Explanations for Relational Deep Learning	2026	VLDB
6	377	TURL: Table Understanding through Representation Learning	2021	VLDB
7	10,241	Scalable and Usable Relational Learning With Automatic Language Bias	2021	SIGMOD
8	2,378	DeepJoin: Joinable Table Discovery with Pre-trained Language Models	2023	VLDB
9	4,631	Transformers for Tabular Data Representation: A Tutorial on Models and Applications	2022	VLDB
10	1,391	Creating Embeddings of Heterogeneous Relational Datasets for Data Integration Tasks	2020	SIGMOD