Hội nghị Quốc tế chuyên đề CNTT lần 6 ngày 3-4/12/2015 tại Huế

ABSTRACT

Privacy   Preserving  Data  Mining  (PPDM)  has  become  an important researctopic in recent years. Hiding high utility sequential patterns isvery necessary in business, health and security applications, etcThgoaohiding is tfind thway to hide all high utility sequential patterns so that the adversaries canomine them from the sanitized database. However, there are a femethodin the literature fohiding high utility sequential patterns. In this paper, wpresent a new approach for solving this problem. First, we use an expansion algorithm of USpan [2] for mining all high utility sequentiapatterns. Then we present two proposed algorithms HHUSP (HidinHigh Utility Sequential Pattern) and MSPCF (Maximum Sequential Patterns Conflict First) to  hide all high utility sequential patterns. Experimental results shothe evaluatioof execution time, memory usagon the large-scale datasets.

 

Dinh Duy Tai, Information Technology Faculty - Hochiminh City College of Industry and Trade;  “A Novel Approach for Hiding High Utility Sequential Patterns”; The 6th international Symposium on Information and Communication Technology 2015 (SoICT 2015) held in Hue City.