Research on Data Approximation Search Privacy Protection in Cloud Environment

In recent years, with the development of technology integration such as the Internet of Things, cloud computing, big data, etc., the volume and speed of data generated annually in the world have increased exponentially. The huge amount of data is hosted on the cloud platform, which aggregates and fuses various data mining algorithms from multiple sources to provide "wisdom" answers to the upper complex application requirements. There is currently no one-dimensional solution for large data search, such as "accuracy, timeliness, privacy granularity". This paper establishes a quantifiable measurement mechanism between the multidimensional indicators with intermediate evaluation criterion "(ϵ,δ)-Approximate" as a bridge, and implements a data retrieval scheme that coordinates the three dimensions of large data search. This mechanism solves the re-search problem caused by isomorphic search and data version update, and improves the efficiency of universal search by improving the Hadoop system architecture.