Facebook contains a large amount of data which can be used for research on interpersonal behavior patterns. However, there has been almost no research on computationally analyzing Facebook data to predict interpersonal behavior. I begin our inquiry into this problem with attachment theory, which conceptualizes how people form relationships with others. In this paper, I seek to reveal how Facebook users` behavior patterns such as status updates and photo albums are related to attachment style. I describe our dataset of 682 Facebook users, their attachment scale survey results, and their 657,057 posts. I define four key features: social capacity, openness, self-disclosure, and emotional disclosure. By analyzing the two dimensions of attachment, anxiousness and avoidance using the four key features, I find significant correlations between the users` attachment style and their Facebook behaviors. The implication is that a computational analysis of a Facebook user`s activities alone can reveal the attachment style of that user without the self-reported survey questionnaire of attachment scale.