Molybdenum palladium photographic inspection in large-scale populations which is called breast cancer screening is a way to effectively reduce the mortality of breast cancer. But in the developing countries or a populous country like China, making molybdenum palladium photographic for breast check age population is not reality, because there will need amount of time, medical resources. For saving cost, time and effort to realize the prevention of breast cancer, reducing the size of the population screening, simultaneously realizing the goal of prevention and control of breast cancer is an effective method. Breast cancer risk assessment model can assess a people is at a high risk of developing the disease breast cancer risk. Consequently, if high-risk people are identified, many unnecessary investigations and procedures can be avoided and medical resources can be saved without compromising medical care. At present, the existing risk assessment models has some limits. In this paper, we put forward a breast cancer risk assessment model based on medical social network (Shrink) which is used to assess the breast cancer risk of people, and further confirm a high-risk group as the screening scope. The model constructs medical social network that expresses the similarity on related risk factors (RRF) of breast cancer between people. The people who have more similarity with breast cancer patients have higher risk. Therefore, according to similarity value, all people in the medical social network lastly are confirmed as high-risk group or low-risk group. The people in high-risk group have higher risk than one in low-risk group, and can be made a further screening. The experiment proves that the assessment efficiency of Shrink is better than Gail model. Also the judgment accuracy proves that judgment function of Shrink is good. Therefore Shrink maybe makes a contribution to the prevention and control of breast cancer.