Abstract: In recent years, the study of argumentation has given birth to a new research field, i.e. computational argumentation. According to the number of people participating in the argumentation process, the research of computational argumentation can be divided into two categories, namely, monological argumentation and dialogical argumentation. The process of dialogical argumentation happens widely in the real world, such as social network platforms, judicial fields, education fields, etc., but related research has just started. This paper reviews the basic task settings, mainstream model frameworks, downstream applications, and public datasets and evaluation methods in the field of dialogical argumentation. Finally, this paper rveals several research directions for the future development, including multi-modal dialogue argumentation analysis, knowledge-infused argumentation generation, etc.