As a massive collection of raw data, data sets frequently breed transaction disputes in practice, giving rise to the demand for property rights confi rmation. However, the special rights model is diffi cult to achieve the self-consistency of data collection rights and industrial prosperity, and the applicability of behavior regulation models has been criticized for its vague and limited protection objects. The similarity between data sets and intellectual property objects establishes a high degree of adaptability between the adjacency rights system and data sets in terms of institutional objectives, protected objects, rights structure, and protection boundaries. Therefore, constructing a system for protecting and restricting the adjacency rights of data sets is currently the institutional choice for achieving the goals of data set production and circulation. In other words, under the framework of copyright law, establish provisions for the protection of adjacent rights of data sets, clarify the protected objects, subjects, and powers of adjacent rights of data sets, and at the same time, construct a system for restricting adjacent rights of data sets based on a balance of interests.