Overview
The HIPE-2026 Shared Task invites participants to engage in the extraction of person-location relations from multilingual historical texts. This third edition builds on the successes of previous iterations in 2020 and 2022, aiming to enhance the analysis of entities and support the reconstruction of individuals’ geographical and temporal paths. The task is significant for the AI and ML community as it addresses the complexities of historical data, which often contains noisy and fragmented information.
Background & Relevance
The extraction of person-location relations is a critical area within natural language processing (NLP) and digital humanities. As researchers seek to understand historical contexts, the ability to accurately link individuals to specific places and times becomes essential. This task not only contributes to advancements in NLP but also aids historians and researchers in uncovering insights from historical documents. The challenge lies in the nuanced interpretation of texts that may not clearly indicate relationships, requiring sophisticated reasoning capabilities from AI systems.
Key Details
- Registration Deadline: April 23, 2026
- Training Data Releases:
- Partial: December 19, 2025
- Full: January 19, 2026
- Evaluation Period: May 5-7, 2026
- Workshop Venue: CLEF Conference, September 21-24, 2026, Jena, Germany
- Evaluation Profiles:
- Accuracy Profile
- Efficiency Profile
- Generalization Profile
- Links:
- HIPE-2026 Website
- Participation Guidelines
- HIPE-2026 Data GitHub Repository
Eligibility & Participation
This opportunity is open to researchers, developers, and students interested in the fields of NLP and digital humanities. Participants are encouraged to develop systems that can effectively analyze historical documents and extract meaningful relationships between persons and locations.
Submission or Application Guidelines
- Register for the HIPE-2026 Shared Task by April 23, 2026.
- Access the training data as it becomes available on the specified dates.
- Develop your system to classify person-location relations based on the provided datasets.
- Submit your participant run by the deadline on May 7, 2026.
- Prepare a notebook paper for submission by May 28, 2026.
More Information
The HIPE shared tasks are part of ongoing efforts to leverage AI technologies for the exploration of historical texts. By focusing on person-location relations, this task aims to foster collaboration between the NLP and digital humanities communities, promoting innovative approaches to historical data analysis. The insights gained from this task could significantly enhance our understanding of historical narratives and the movements of individuals through time and space.
Conclusion
The HIPE-2026 Shared Task presents a unique opportunity for participants to contribute to the evolving field of NLP and historical analysis. Researchers are encouraged to explore this challenge, develop innovative solutions, and share their findings with the community. For more information, visit the HIPE-2026 website and consider registering to participate in this exciting initiative.
Category: CFP & Deadlines
Tags: nlp, historical texts, person-location relations, multilingual processing, clef, hipe, entity recognition, temporal reasoning, geographical inference, language models
