WIKITIDE: A Wikipedia-Based Timestamped Definition Pairs Dataset
Borkakoty, Hsuvas, Espinosa-Anke, Luis
–arXiv.org Artificial Intelligence
A fundamental challenge in the current NLP context, dominated by language models, comes from the inflexibility of current architectures to 'learn' new information. While model-centric solutions like continual learning or parameter-efficient fine tuning are available, the question still remains of how to reliably identify changes in language or in the world. In this paper, we propose WikiTiDe, a dataset derived from pairs of timestamped definitions extracted from Wikipedia. We argue that such resource can be helpful for accelerating diachronic NLP, specifically, for training models able to scan knowledge resources for core updates concerning a concept, an event, or a named entity. Our proposed end-to-end method is fully automatic, and leverages a bootstrapping algorithm for gradually creating a high-quality dataset. Our results suggest that bootstrapping the seed version of WikiTiDe leads to better fine-tuned models. We also leverage fine-tuned models in a number of downstream tasks, showing promising results with respect to competitive baselines.
arXiv.org Artificial Intelligence
Aug-18-2023
- Country:
- South America > Argentina
- Pampas > Buenos Aires F.D. > Buenos Aires (0.04)
- Oceania > Australia
- North America
- Canada > Quebec (0.14)
- United States
- Pennsylvania (0.04)
- New York (0.04)
- Massachusetts (0.04)
- Texas > Travis County
- Austin (0.04)
- Minnesota > Hennepin County
- Minneapolis (0.14)
- Illinois > Cook County
- Chicago (0.04)
- Europe
- Germany > Berlin (0.04)
- Belgium > Wallonia (0.04)
- Sweden > Östergötland County
- Linköping (0.04)
- Ireland > Leinster
- County Dublin > Dublin (0.04)
- Finland > Southwest Finland
- Turku (0.04)
- Asia > Thailand
- South America > Argentina
- Genre:
- Research Report > New Finding (0.68)
- Industry:
- Leisure & Entertainment > Sports > Football (1.00)
- Technology: