DBScholar

Back to papers

Towards Estimation Error Guarantees for Distinct Values

Summary: Prove any sublinear-sample estimator for distinct counts must suffer large error on some natural distributions unless it reads a large fraction of the data. Give an estimator matching this lower bound and practical heuristics for typical distributions, validated empirically. (summarized by gpt-5-mini on Feb 09 2026)

Paper ID
1216
Venue
PODS
Year
2000
Pagerank
0.00022296371
Overall Rank
288 | 98.03%
DOI
10.1145/335168.335230

Incoming Non-self Citations Over Time

Authors

BibTeX Citation

@inproceedings{charikar_pods00,
        address = {New York, NY, USA},
        series = {{PODS} '00},
        title = {{Towards Estimation Error Guarantees for Distinct Values}},
        url = {https://dl.acm.org/doi/10.1145/335168.335230},
        doi = {10.1145/335168.335230},
        booktitle = {Proceedings of the {ACM} {SIGMOD} Symposium on {Principles} of {Database} {Systems}},
        publisher = {Association for Computing Machinery},
        author = {Charikar, Moses and Chaudhuri, Surajit and Motwani, Rajeev and Narasayya, Vivek},
        year = {2000}
}

Incoming Citations (Sorted by Pagerank)

Showing 9 of 59 citing papers.

Previous Page 2 / 2 Next

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.

Previous Page 1 / 1 Next

Semantically Similar Papers