Data Mining and Knowledge Engineering

B. Hathi Karishma ¹, P. Desai Sonali ¹, A. Varsur Jalpa ¹

Affiliations
1 CSE Department, B.H. Gardi College of Engg. & Tech., Rajkot, Gujarat, IN

Abstract

Sequential pattern mining is a very important mining technique with wide applications. It found very useful in various domains like natural disaster, sales record analysis, marketing strategy, shopping sequences, medical treatment and DNA sequences etc. It discovers the subsequence's and frequent relevant pattern from the given sequences in the database. The first was Apriori algorithm, which was put forward by the founders themselves. Later more scalable algorithms for difficult applications were developed. E.g. GSP, Spade, PrefixSpan etc. In this paper, a survey of the sequential pattern mining algorithms is given.

Keywords

Sequential Pattern, Sequence Database, Itemsets, Apriori, Pattern Growth.

Full Text

     

R. Manvar Sagar ¹, P. Desai Sonali ¹, B. Hathi Karishma ¹

Affiliations
1 CE Department, B.H. Gardi College of Engg. & Tech., Rajkot, Gujarat, IN

Abstract

The advancement in data mining techniques plays an important role in many applications. In context of privacy and security issues, the problems caused by association rule mining technique are investigated by many research scholars. It is proved that the misuse of this technique may reveal the database owner's sensitive and private information to others. Many researchers have put their effort to preserve privacy in Association Rule Mining. Amongst the two basic approaches for privacy preserving data mining, viz. Randomization based and Cryptography based, the later provides high level of privacy but incurs higher computational as well as communication overhead. Hence, it is necessary to explore alternative techniques that improve the over-heads. In this work, we propose an efficient, collusion-resistant cryptography based approach for distributed Association Rule mining using Shamir's secret sharing scheme. As we show from theoretical and practical analysis, our approach is provably secure and require only one time a trusted third party. We use secret sharing for privately sharing the information and code based identification scheme to add support against malicious adversaries.

Keywords

Privacy, Privacy Preservation in Data Mining (PPDM), Horizontally Partitioned Database, EMHS, MFI, Shamir Secret Sharing.

Full Text