抽象的な

Term Frequency Based Sequence Generation Algorithm for Graph Based Data Anonymization

S.Charanyaa, K.Sangeetha

A big deal of research has been performed in the area of graphical data anonymization. Because of the wide range of application of graphical data from social network data to large warehouse data and knowledge engineering domains. Notion of k-anonymity has been proposed in literature, which is a framework for protecting privacy, emphasizing the lemma that a database to be k-anonymous, every tuple should be different from at least k-1 other tuples in accordance with their quasi-identifiers(QIDs). Inspite of the existence of k-anonymity framework, malicious users and misfeasers may get authorization to the sensitive information if a set of nodes exhibit alike attributes. In this paper we make a systematic analysis on structure anonymization mechanisms and models projected in the literature. Also we discuss the simulation analysis of KDLD model creation and construction. We propose a Term Frequency Based Sequence Generation Algorithm (TFSGA) which creates node sequence based on term frequency of tuples with minimal distortion. We experimentally show the efficiency of the proposed algorithm under varying cluster sizes.