Assessing Political Prudence of Open-domain Chatbots
This work addresses the problem of ensuring responsible and engaging behavior in chatbots for users and developers, but it is incremental as it focuses on assessment rather than solving the underlying issue.
The authors tackled the challenge of politically sensitive topics in open-domain chatbots by proposing a set of metrics to assess political prudence, analyzing various chatbots through automatic and human evaluations, and releasing testsets and code to advance research.
Politically sensitive topics are still a challenge for open-domain chatbots. However, dealing with politically sensitive content in a responsible, non-partisan, and safe behavior way is integral for these chatbots. Currently, the main approach to handling political sensitivity is by simply changing such a topic when it is detected. This is safe but evasive and results in a chatbot that is less engaging. In this work, as a first step towards a politically safe chatbot, we propose a group of metrics for assessing their political prudence. We then conduct political prudence analysis of various chatbots and discuss their behavior from multiple angles through our automatic metric and human evaluation metrics. The testsets and codebase are released to promote research in this area.