Mining an "Anti-Knowledge Base" from Wikipedia Updates with Applications to Fact Checking and Beyond

Summary: Introduces unsupervised anti-knowledge mining: detecting Wikipedia corrections, estimating correction likelihood via iterative EM over Web-claim frequencies, and extracting ranked erroneous SPO triples. Produces 110K long-tail mistakes at high precision, enabling fact-checking benchmarks and Web-wide error discovery. (summarized by gpt-5.6-luna on Jul 24 2026)

Paper ID: h34a254f26483de74
Venue: VLDB
Year: 2020
Pagerank: 6.0404239e-05
Overall Rank: 5,679 | 61.82%
DOI: 10.14778/3372716.3372727

Incoming Non-self Citations Over Time

Authors

1. Georgios Karagiannis (Cornell University)
2. Immanuel Trummer (Cornell University)
3. Saehan Jo (Cornell University)
4. Shubham Khandelwal (Cornell University)
5. Xuezhi Wang (Google)
6. Cong Yu (Google)

BibTeX Citation

@article{karagiannis_vldb20,
        title = {{Mining an "Anti-Knowledge Base" from Wikipedia Updates with Applications to Fact Checking and Beyond}},
        author = {Karagiannis, Georgios and Trummer, Immanuel and Jo, Saehan and Khandelwal, Shubham and Wang, Xuezhi and Yu, Cong},
        journal = {PVLDB},
        series = {{VLDB} '20},
        volume = {13},
        number = {4},
        pages = {561--574},
        doi = {10.14778/3372716.3372727},
        url = {https://doi.org/10.14778/3372716.3372727},
        year = {2020}
}

Incoming Citations (Sorted by Pagerank)

Showing 5 of 5 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
11,149	Demonstrating CEDAR: A System for Cost-Efficient Data-Driven Claim Verification	2025	SIGMOD	4.9793485e-05
11,364	CEDAR: A System for Cost-Efficient Data-Driven Claim Verification	2025	VLDB	4.9793485e-05
11,970	To Intervene or Not To Intervene: Cost based Intervention for Combating Fake News	2021	SIGMOD	4.9793485e-05
12,021	Wikinegata: a Knowledge Base with Interesting Negative Statements	2021	VLDB	4.9793485e-05
12,034	On the Limits of Machine Knowledge: Completeness, Recall and Negation in Web-scale Knowledge Bases	2021	VLDB	4.9793485e-05

Outgoing Citations (Sorted by Pagerank)

Showing 9 of 9 cited papers.

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

Rank	Cited Paper	Year	Venue	Pagerank
104	HoloClean: Holistic Data Repairs with Probabilistic Inference	2017	VLDB	0.00033690989
205	Snorkel: Rapid Training Data Creation with Weak Supervision	2018	VLDB	0.00025181304
328	Declarative Information Extraction Using Datalog with Embedded Extraction Predicates	2007	VLDB	0.00020928928
2,272	Resolving Conflicts in Heterogeneous Data by Truth Discovery and Source Reliability Estimation	2014	SIGMOD	8.7121962e-05
2,482	ClaimBuster: The First-ever End-to-end Fact-checking System	2017	VLDB	8.399913e-05
3,171	Truth Inference in Crowdsourcing: Is the Problem Solved?	2017	VLDB	7.5691557e-05
5,138	Verifying Text Summaries of Relational Data Sets	2019	SIGMOD	6.2579762e-05
6,428	Domain-Aware Multi-Truth Discovery from Conflicting Sources	2018	VLDB	5.7886234e-05
7,203	Mining Subjective Properties on the Web	2015	SIGMOD	5.5865623e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	12,021	Wikinegata: a Knowledge Base with Interesting Negative Statements	2021	VLDB
2	7,274	Computational Fact Checking: A Content Management Perspective	2018	VLDB
3	9,384	Selecting Data to Clean for Fact Checking: Minimizing Uncertainty vs. Maximizing Surprise	2019	VLDB
4	9,460	Combating Fake News: A Data Management and Mining Perspective	2019	VLDB
5	2,960	Auto-Detect: Data-Driven Error Detection in Tables	2018	SIGMOD
6	12,278	Building Structured Databases of Factual Knowledge from Massive Text Corpora	2017	SIGMOD
7	9,494	Automatically Generating Interesting Facts from Wikipedia Tables	2019	SIGMOD
8	3,552	Knowledge-Based Trust: Estimating the Trustworthiness of Web Sources	2015	VLDB
9	3,178	Toward Computational Fact-Checking	2014	VLDB
10	7,778	User Guidance for Efficient Fact Checking	2019	VLDB