Collective behavior through interactions is ubiquitous in living systems ranging from cells to animal groups, and leads to rapid and accurate decision-making or efficient division of labor. Especially, it has shown that social insects such as ants and bees exhibit sophisticated collective behavior, which has drawn much attention in behavioral ecology. However, how the network structure and information flow can emerge from local interactions is less understood. Here, we recorded the trajectory of movements in ants (Diacamma sp.) using a tag-based tracking system, and detected interactions between individuals using two ways: time-series analysis and proximity of individuals. We show the interaction patterns and untangled collective dynamics, and then discuss information flows in ant society.