Robust Generalizable Heterogeneous Legal Link Prediction
Lorenz Wendlinger0000-0001-9459-6244 1, Simon Alexander Nonn0009-0007-7173-5302 1, Abdullah Al Zubaer0009-0001-0842-7434 1, Michael Granitzer0000-0003-3566-5507 1,2 · 2026
Read Robust Generalizable Heterogeneous Legal Link Prediction by Lorenz Wendlinger0000-0001-9459-6244 1, Simon Alexander Nonn0009-0007-7173-5302 1, Abdullah Al Zubaer0009-0001-0842-7434 1, Michael Granitzer0000-0003-3566-5507 1,2 (2026) 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-2602-04812.txt.
Description
Lorenz Wendlinger and 3 co-authors improve link prediction on legal citation networks in a 2026 study. These networks mix statutes, cases, and rich metadata, and predicting missing links can surface relevant law. Recent results, the authors find, leave room for stronger and more portable models.
Two changes drive their gains. Edge dropout in training and joining features yield more robust representations, cutting error rates by a lot. A second approach uses multilingual node features with an improved asymmetric decoder built for fit between varied node types. That version extends prediction to data from New Zealand, far apart in place and language from the sources. The changes also improved transfer between disjoint legal systems.
Catalog record
| CREATOR | Lorenz Wendlinger0000-0001-9459-6244 1, Simon Alexander Nonn0009-0007-7173-5302 1, Abdullah Al Zubaer0009-0001-0842-7434 1, Michael Granitzer0000-0003-3566-5507 1,2 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-04812