Abstract
Community mining aims to find groups with dense relations among its members from networks such as those on SNSs. It is used not only in data mining and clustering but also for directly obtaining knowledge from the groups. Community mining in researcher networks has been studied for a long time, and is recognized as a method of capturing the topics and activities in research areas. We are studying community mining in terms of networks given by relations between researchers, such that two researchers meeting at a conference are considered to be related, whereas previous studies have used co-author graphs that are defined by the relation of having jointly written papers. We examined our idea of using a network constructed from a database of the society IEICE, which includes information on papers submitted to domestic workshops. As a result, we were able to observe many interesting communities such as those composed of people attending several types of workshops.
