Organization structure analysis based on an affiliation network, and verification of its effectiveness

Groups in an organization are not always statically affiliated with organizations in the nominal organizational structure but change dynamically according to the task or the project the groups participate in. Therefore, it is difficult to understand the real interactions within and between groups. This paper aims to identify groups in their real states by using the affiliation network concept. It uses email header information to identify virtual groups on the basis of cliques and clans. A case study at an existing research organization has been conducted to verify the effectiveness of our approach. The results show that the method is effective in identifying true groups, both formal and informal. This approach also enables the identification of leadership roles within the groups. An analysis of mail data collected over a two-month span in 2006 was compared with that for data one year later in 2007, and showed that the proposed method can be effective in identifying changes in the real, effective organization. These analyzed results also show that the human network in this organization has the tendency to be the small world in proportion to the activity in the organization.