RuleBert: Teaching Soft Rules to Pre-trained Language Models
Saeed, Mohammed, Ahmadi, Naser, Nakov, Preslav, Papotti, Paolo
–arXiv.org Artificial Intelligence
While pre-trained language models (PLMs) are the go-to solution to tackle many natural language processing problems, they are still very limited in their ability to capture and to use common-sense knowledge. In fact, even if information is available in the form of approximate (soft) logical rules, it is not clear how to transfer it to a PLM in order to improve its performance for deductive reasoning tasks. Here, we aim to bridge this gap by teaching PLMs how to reason with soft Horn rules. We introduce a classification task where, given facts and soft rules, the PLM should return a prediction with a probability for a given hypothesis. We release the first dataset for this task, and we propose a revised loss function that enables the PLM to learn how to predict precise probabilities for the task. Our evaluation results show that the resulting fine-tuned models achieve very high performance, even on logical rules that were unseen at training. Moreover, we demonstrate that logical notions expressed by the rules are transferred to the fine-tuned model, yielding state-of-the-art results on external datasets.
arXiv.org Artificial Intelligence
Sep-24-2021
- Country:
- North America
- Puerto Rico (0.04)
- United States
- Minnesota > Hennepin County
- Minneapolis (0.14)
- Louisiana > Orleans Parish
- New Orleans (0.04)
- California
- San Francisco County > San Francisco (0.14)
- Monterey County > Monterey (0.04)
- Los Angeles County > Long Beach (0.04)
- Minnesota > Hennepin County
- Canada
- Europe
- Netherlands (0.04)
- Austria (0.04)
- United Kingdom > England
- Cambridgeshire > Cambridge (0.04)
- Spain > Catalonia
- Barcelona Province > Barcelona (0.04)
- Italy > Tuscany
- Florence (0.04)
- France > Île-de-France
- Asia
- Middle East > Qatar (0.04)
- China > Hong Kong (0.04)
- Africa
- South Africa > Western Cape
- Cape Town (0.04)
- Ethiopia > Addis Ababa
- Addis Ababa (0.04)
- South Africa > Western Cape
- North America
- Genre:
- Research Report (0.70)
- Technology: