In recent years, there have been increasing opportunities for discussion and exchange of opinions on the internet. In order to clearly understand such discussions, users should be able to fully understand the flow of discussion as well as the conclusions. It is especially important to understand the main points as well as the trends in people's opinions in order to have more than a superficial understanding of the discussion. However, it is often difficult to understand them as the number of comments increases over time and the discussions become complex. In this research, we aim to develop methods for visualizing discussions in comments on online news articles to understand opinion trends and discussion points. This paper describes the methods for generating summaries of discussions and presents an overview of the support system. We also discuss the features of the proposed methods through the results of experiments using a prototype system for visualizing discussions.