Community detection is an important line of research in understanding complex networks. In many real-world networks, communities naturally overlap since a node usually has multiple community memberships, which makes overlapping community detection a trend in recent years. One popular technique to cope with overlapping community detection is matrix factorization (MF). In this book, we review existing models in matrix factorization framework for overlapping community detection and propose four novel models to explore the relationship between links and communities. By evaluate our proposed models on several real-world networks including large ones with ground-truth communities, we show that our models outperform state-of-the-art approaches on multiple measurements and are capable of finding overlapping communities with better quality. This book should be especially useful to researchers focusing on community detection, or anyone else who is interested in social networks.