CLEF HIPE-2026: Evaluating Accurate and Efficient Person-Place Relation Extraction from Multilingual Historical Texts
Juri Opitz, Corina Raclé, Emanuela Boros, Andrianos Michail, Matteo Romanello, Maud Ehrmann, Simon Clematide
TL;DR
HIPE-2026 tackles the problem of extracting person–place relations from multilingual historical texts by defining two relation types, $at$ and $isAt$, and framing the task with a temporal horizon relative to publication. It introduces three evaluation profiles—accuracy, efficiency, and generalization—and provides multilingual data sets (Test Set A and Surprise Set B) along with baselines and tooling to support robust, domain-general RE. The work addresses gaps in multilingual, noisy, and domain-shift relation extraction within historical corpora and highlights abductive reasoning as a core interpretive strategy. By enabling reliable spatio-temporal reconstruction, HIPE-2026 supports knowledge-graph construction and digital humanities analyses across literature and press archives.
Abstract
HIPE-2026 is a CLEF evaluation lab dedicated to person-place relation extraction from noisy, multilingual historical texts. Building on the HIPE-2020 and HIPE-2022 campaigns, it extends the series toward semantic relation extraction by targeting the task of identifying person--place associations in multiple languages and time periods. Systems are asked to classify relations of two types - $at$ ("Has the person ever been at this place?") and $isAt$ ("Is the person located at this place around publication time?") - requiring reasoning over temporal and geographical cues. The lab introduces a three-fold evaluation profile that jointly assesses accuracy, computational efficiency, and domain generalization. By linking relation extraction to large-scale historical data processing, HIPE-2026 aims to support downstream applications in knowledge-graph construction, historical biography reconstruction, and spatial analysis in digital humanities.