More number of users requires cloud services to transfer private data like electronic health records and financial transaction records. A cloud computing services offers several flavors of virtual machines to handle large scale datasets. But centralized approaches are difficult in handling of large datasets. Data anonymization is used for privacy preservation techniques. It is challenged to manage and process such large-scale data within a cloud application. A scalable two-phase top-down specialization (TDS) approach to anonymize large-scale data sets using the Map Reduce framework on cloud. It is used to investigate the scalability problem of large-scale data anonymization techniques. These approaches deliberately design a group of innovative Map Reduce jobs to concretely accomplish the specialization computation in a highly scalable way. The Top-Down Specialization process speeds up the specialization process because indexing structure avoids frequently scanning entire data sets and storing statistical results.