Integrity Verification of K-means Clustering Outsourced to Infrastructure as a Service (IaaS) Providers

Abstract

The Cloud-based infrastructure-as-a-service (IaaS) paradigm (e.g., Amazon EC2) enables a client who lacks computational resources to outsource her dataset and data mining tasks to the Cloud. However, as the Cloud may not be fully trusted, it raises serious concerns about the integrity of the mining results returned by the Cloud. To this end, in this paper, we provide a focused study about how to perform integrity verification of the k-means clustering task outsourced to an IaaS provider. We consider the untrusted sloppy IaaS service provider that intends to return wrong clustering results by terminating the iterations early to save computational cost. We develop both probabilistic and deterministic verification methods to catch the incorrect clustering result by the service provider. The deterministic method returns 100% integrity guarantee with cost that is much cheaper than executing k-means clustering locally, while the probabilistic method returns a probabilistic integrity guarantee with computational cost even cheaper than the deterministic approach. Our experimental results show that our verification methods can effectively and efficiently capture the sloppy service provider.

Result Integrity Verification of Outsourced Privacy-preserving Frequent Itemset Mining

“…Unfortunately, this body of theory is impractical [19], and thus it is difficult to adopt these general-purpose cryptographic verification techniques to data mining problems for practical verification. Recently several result integrity verification methods have been designed for specific data mining problems, including frequent itemset mining [20,7,8], outlier mining [17], clustering [15]. Among these work, only [20,7,8] focus on integrity verification methods for frequent itemset mining.…”

Result Integrity Verification of Outsourced Bayesian Network Structure Learning

“…Only a few work [18,11,12] have studied the issue of integrity verification of data mining computations outsourced to third-party DM aS providers. [18] studies the problem of verifying the correctness and completeness of outsourced association rule mining.…”

“…Its basic idea is to insert artificial items into the dataset; the mining result can be verified via the artificial (in)frequent itemsets constructed from artificial items. [11] and [12] apply the same strategy of using artificial data mining objects to the problems of verifying outlier mining and k-means clustering respectively. These work show that the design of artificial mining objects vary dramatically for different data mining problems.…”

“…However, it is not clear how to apply these verification techniques to data mining problems, due to the complexity of these techniques and the difficulty of using general-purpose cryptographic techniques to solve specific data mining problems. Some recent research work [18,11,12] designed result integrity assurance techniques for various data mining problems, including association rule mining [18], outlier mining [11], and k-means clustering [12]. These work show that the design of verification methods can vary significantly for different data mining problems and applications.…”

Protect Both Integrity and Confidentiality in Outsourcing Collaborative Filtering Computations

“…Besides [9] and [10], there are many other related works. For example, Wong et al [11] and Dong et al [4] investigated the integrity issues in outsourcing frequent itemset mining computations, and Liu et al [6] proposed probabilistic and deterministic methods to verify clustering results for k-means clustering algorithms.…”