RENOVI: A Benchmark Towards Remediating Norm Violations in Socio-Cultural Conversations
This addresses the need for culturally-aware AI in conversational systems, though it's incremental as it primarily provides a dataset and benchmark.
The authors tackled the problem of AI systems lacking the ability to remediate norm violations in socio-cultural conversations by creating ReNoVi, a benchmark dataset of 9,258 multi-turn dialogues (512 human-authored and 8,746 synthetic) with annotations for social norms, which showed performance improvements from using synthetic data.
Norm violations occur when individuals fail to conform to culturally accepted behaviors, which may lead to potential conflicts. Remediating norm violations requires social awareness and cultural sensitivity of the nuances at play. To equip interactive AI systems with a remediation ability, we offer ReNoVi - a large-scale corpus of 9,258 multi-turn dialogues annotated with social norms, as well as define a sequence of tasks to help understand and remediate norm violations step by step. ReNoVi consists of two parts: 512 human-authored dialogues (real data), and 8,746 synthetic conversations generated by ChatGPT through prompt learning. While collecting sufficient human-authored data is costly, synthetic conversations provide suitable amounts of data to help mitigate the scarcity of training data, as well as the chance to assess the alignment between LLMs and humans in the awareness of social norms. We thus harness the power of ChatGPT to generate synthetic training data for our task. To ensure the quality of both human-authored and synthetic data, we follow a quality control protocol during data collection. Our experimental results demonstrate the importance of remediating norm violations in socio-cultural conversations, as well as the improvement in performance obtained from synthetic data.