This paper proposes a technique for generating more comprehensible animations from discussions, which are often hard to follow, in USENET. This technique consists of two steps. In the first step, our prototype system generates a scenario from articles in a news thread using the quote relationship. In the second step, it generates an animation based on the scenario, casting 3D avatars as the authors of the articles. We also implemented a prototype system based on this technique and made several animations from articles posted to USENET. Keywords Automatic Animation, USENET, using 3D Avatar