ScholarWorks@Y-Scholar Hub

초록

Understanding on the identity of the online community Megalia, which adopted hostile activism strategy against sexism, is diversified from as a representation of a new generation’s feminism to as simply a hate-filled website. This study explores reasons for the divergent interpretations on the identity of Megalia, by adopting and modifying the social identity model of deindividuation effects (SIDE) for anonymous online communities. We analyzed the entire 162,893 posts throughout the lifecourse of Megalia from 2015 to 2016, in order to figure out how discourses were differentiated depending on whether the writer's ID was disclosed (pseudonym) or not (anonym), and to reconstruct the evolvement of collisions between the discourses. Natural language processing techniques (i.e., topic modeling and Word2Vec) were applied. Then, it discovered that topics related to gender discrimination were dominant whereas topics related to internal conflicts were relatively minor. Anonymous-posts, however, exclusively preferred anger-raising topics with Megalian dialect while pseudonymous-posts contained topics related to mobilizing efforts against sexual crimes and discriminations. An observer would see very different faces of Megalia, depending on whether reading anonymous posts, which made up most of posts, or following significant contributors' IDs and their pseudonymous posts. In addition, even a shared discourse between the two types of identities may nurture a conflict by developing either type of termonology exclusive to each identity or differential contexts of a common word. A topic on tolerance toward gender minorities developed both patterns and eventually contributed to massive exits and final decline. We expect to contribute to the development of analytical frameworks that clarify how members in an online community form various identities in response to structural conditions such as levels of anonymity and to foster holistics understanding of the life-course of the community.