Disinformation: Detection or Treatment?

Nowadays, most people have realized the impact of disinformation on our daily life. Many AI technologies have been applied to deal with it, especially for detecting it. However, users may have read the disinformation and believed it before the AI models can detect and remove it. In this talk, I will introduce some of the possible further solutions for changing people’s belief on the fake information. We focus on the way to provide the corrected information and explanations to users. We hope these methods can cover more target users and will be more effective as a kind of “treatment” rather than pure “diagnosis”.

當大家對假新聞的議題都能評論幾句的今日，除了使用 AI 技術自動找出假新聞之外，我們還能做甚麼呢？當假新聞在我們的日常生活氾濫成災，哪些人是接觸假新聞的高危險群呢？已經閱讀假新聞的使用者，我們該如何幫助他們？如何告訴大眾為什麼這是假新聞？他們會相信嗎？如何更好地解釋新聞之所以不真實的原因並扭轉讀者的認知？人工智慧技術在假新聞的議題上，除了偵測，我們能做的還有更多。在這個演講中，我將介紹假新聞偵測下一步的技術，並探討以人工智慧處理假新聞可能遇到的問題。