Foveate, Attribute, and Rationalize: Towards Physically Safe and Trustworthy AI
Preprint
Abstract
Users' physical safety is an increasing concern as the market for intelligent systems continues to grow, where unconstrained systems may recommend users dangerous actions that can lead to serious injury. Covertly unsafe text is an area of particular interest, as such text may arise from everyday scenarios and are challenging to detect as harmful. We propose FARM, a novel framework leveraging external knowledge for trustworthy rationale generation in the context of safety. In particular, FARM foveates on missing knowledge to qualify the information required to reason in specific scenarios and retrieves this information with attribution to trustworthy sources. This knowledge is used to both classify the safety of the original text and generate human-interpretable rationales, shedding light on the risk of systems to specific user groups and helping both stakeholders manage the risks of their systems and policymakers to provide concrete safeguards for consumer safety. Our experiments show that FARM obtains state-of-the-art results on the SafeText dataset, showing absolute improvement in safety classification accuracy by 5.9%.Keywords
All Related Versions
This publication has 23 references indexed in Scilit:
- Drink Bleach or Do What Now? COVID-HeRA: A Study of Risk-Informed Health Decision Making in the Presence of COVID-19 MisinformationProceedings of the International AAAI Conference on Web and Social Media, 2022
- Risk-graded Safety for Handling Medical Queries in Conversational AI2022
- Generated Knowledge Prompting for Commonsense Reasoning2022
- Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLPTransactions of the Association for Computational Linguistics, 2021
- Confronting Abusive Language Online: A Survey from the Ethical and Human Rights PerspectiveJournal of Artificial Intelligence Research, 2021
- Zero-shot Fact Verification by Claim Generation2021
- GoldenWind at SemEval-2021 Task 5: Orthrus - An Ensemble Approach to Identify Toxicity2021
- Detecting Cyberbullying and Cyberaggression in Social MediaACM Transactions on the Web, 2019
- A Knowledge-Grounded Neural Conversation ModelProceedings of the AAAI Conference on Artificial Intelligence, 2018
- Approaches to Automated Detection of Cyberbullying: A SurveyIEEE Transactions on Affective Computing, 2017