| 講題：Privacy-Preserving Outsourcing of Frequent Itemset Mining 主講人：戴志華 博士(國立臺灣大學電機工程研究所博士) 時間：101年3月14日 (星期三13:30) 地點：三峽校區社科大樓社1F06教室 Abstract Data mining is the analyzing technique that can discover implicit information from very large databases. Recent years data mining has shown its great promise in wide fields, such as business, bioinformatics, location-based services, and networking analysis, to name a few. Nowadays, many big companies, such as Amazon, Google, IBM and Microsoft, provide data mining services with cloud computing. For everyone who lacks of expertise in data mining and/or computational resources, this allows her/him to be able to take the advantages of data mining by outsourcing. In this talk, I will focus on the outsourcing of frequent itemset mining, which is to discover what usually happens together in large data sets. As privacy is always a concern in outsourcing, I will first discuss the risks and challenges of outsourcing frequent itemset mining, and present a secured model for this problem afterward. The model can limit the probability of privacy leaks with moderate storage overhead. |
