Correct decision-making and effective implementation of public policies will bring positive results to the development of the economy and the advancement of society. Errors in decision-making or improper implementation will bring certain negative effects. In a sense, the issue of public policy is the most important issue of a country's legislation and judiciary, and it is also the main method used by the authorities to safeguard the interests of the people. With the development of the Internet, a lot of information can be effectively processed through computer network technology, so information feedback has become the goal that people pursue, and the online voting system is one of the products. In this paper, a network online voting system is specially designed for public policy formulation and decision-making. The association rules algorithm in the data mining algorithm is introduced, and the system can exclude the user IPs who voted repeatedly. Through the algorithm statistical data results, the voting rate and voting items can be found. The algorithm also optimizes the statistical process of voting results, making voting data processing more convenient and accurate. After the system is tested, the operational reliability of the system in this paper is also verified.

The inherent decentralization and security properties of blockchain make it well suited to address today's educational decision-making challenges. The educational policy evaluation index system is built on the basis of blockchain technology. The purpose of this paper is to study and implement educational policy formulation based on blockchain technology, and to systematically analyze the theoretical basis for educational policy formulation and evaluation using blockchain technology-related theories, combined with normative research, comparative research, and empirical research methods. This paper analyzes the process of educational policy formulation, designs an educational policy evaluation model, and provides a systematic and standardized management method for educational policy formulation and evaluation. The indicator weight calculation results show that the policy effect accounts for 43%, the policy effect accounts for 23%, and the policy effect accounts for 25%. In addition, taking the education policy of M province as an example, to verify the scientificity and operability of the education policy evaluation model established in this paper, the results show that the actual effective rate of the province's education policy is 63.56%, and the overall result is "average". It belongs to the middle and lower classes, which is in line with the actual situation.