Understanding the Effect of Knowledge Graph Extraction Error on Downstream Graph Analyses: A Case Study on Affiliation Graphs
Erica Cai, Brendan O’Connor · 2025
Read Understanding the Effect of Knowledge Graph Extraction Error on Downstream Graph Analyses: A Case Study on Affiliation Graphs by Erica Cai, Brendan O’Connor (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-12367.txt.
Description
Erica Cai and Brendan O'Connor ask in a 2025 paper how errors in machine built knowledge graphs skew later study. Such graphs map who belongs to which group and aid work from social science to health. Large models now pull these graphs from huge text sets, but the cost of their slips is unclear.
The study checks errors at two levels, single links and whole graph shape, on graphs of person membership pulled from social register books. There is a band of pull quality where bias in most graph measures stays near zero. Below that band, bias grows and each measure drifts in its own steady way. Fake error tests show common error models miss that drift, which warns users to check quality first.
Catalog record
| CREATOR | Erica Cai, Brendan O’Connor |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-12367