Analysis and Improvement of Privacy-Preserving Frequent Item Protocol for Accountable Computation Framework

Nowadays, data collection and processing becomes ubiquitous in social and business areas, especially in Internet of Things. However, sensitive information leakage is a critical issue. To solve problem, privacy-preserving techniques are strongly needed. Jiang et al. proposed a protocol of finding frequent item in accountable computing (AC) framework which enables two parties to conduct collaborative computation on their transactional databases to find out the common frequent items without disclosing their private data to the other party. Their scheme was proposed in a secure two-party computation model against malicious adversaries. In this paper, we analyze the implementation details of AC-framework and identify some security weaknesses in their scheme. Furthermore, we clarify the security requirements for the AC-framework and present an augmented solution to enhance security.