The powerful resource advantage of the cloud provides a suitable computing environment for data processing. By transferring local computing to the cloud, the efficiency of data processing can be improved. However, the open cloud environment has defects in data privacy-preserving. In order to strengthen the protection of data privacy and ensure the security of multi-party interaction, we propose a privacy-preserving multi-party computing scheme for K-means clustering (PPMCK). PPMCK can preserve data privacy in the cloud and in the local side for each party from multi-party computing. In addition, PPMCK uses homomorphic encryption to protect data privacy. To support the division operation and ciphertext value size comparison with which homomorphic encryption cannot handle, the corresponding measurements are adopted, which make homomorphic encryption work smoothly. The experimental results demonstrate that PPMCK is effective in both data processing and privacy-preserving.

云的强大资源优势为数据处理提供了合适的计算环境。通过将本地计算转移到云中，可以提高数据处理的效率。但是，开放云环境在保存数据隐私方面存在缺陷。为了加强对数据隐私的保护并确保多方交互的安全性，我们提出了一种用于K均值聚类（PPMCK）的保护隐私的多方计算方案。PPMCK可以为多方计算中的每一方保留云中和本地方的数据隐私。此外，PPMCK使用同态加密来保护数据隐私。为了支持同态加密无法处理的除法运算和密文值大小比较，采用了相应的度量，可以使同态加密顺利进行。实验结果表明，PPMCK在数据处理和隐私保护方面均有效。