This paper addresses an important and interesting problem-automatically identifying adult accounts on Sina Weibo. The authors propose two sets of behavior indicators for adult groups and accounts, and find that adult groups and accounts have different behavioral distributions with non-adult groups and accounts. Then a novel relation-based model, which considers the inter-relationships among groups, individual accounts and message sources, is applied to identify adult accounts. The experimental results show that compared with state-of-the-art methods, the proposed method can improve the performance of adult account identification on Sina Weibo.

Overall, the article is well organized and its presentation is good. However, some minor issues still need to be improved:

(1) The authors should summarize the main contributions of this paper in Section 1.

(2) In Section 4.2, the authors mentioned that “A group will attain a value very close to on GACS if

all its accounts have entirely copied their own texts, images or contact information”. However, according to Equation 8, contact information is not considered when computing GACS. (3) In Algorithm 1 on Pg. 17, it seems that “t=t+1” should be added after line 6. (4) I suggest that the limitation of this work should be discussed in Section 9. (5) There are a few typos and grammar errors in this paper.