主讲人简介：
支志雄，现任澳大利亚联邦科学与工业组织（CSIRO）Data61研究所，云计算和传感数据安全组的科研组长。在加入CSIRO之前，他从普渡大学获得博士学位，先后在Philips Research、IBM Poughkeepsie、中国香港大学、新加坡国立大学、清华大学任职，并担任WCW 2004, AWCC 2004, IEEE SOSE 2006, ICSOC 2009, SCC 2009, SIE 2010, EDOC 2011, EDOC 2012, CWI 2012, SCC 2014, ICICS2016, and ICBE 2016等会议主席，已在国际知名会议和期刊上发表论文超过250篇，拥有6项已经产业化的美国专利。他目前的研究领域包括行为信息学和分析学，网络安全，物联网，云计算、服务计算和社交网络。
报告摘要：
Abstract: Under the era of Big Data, people have been exploring ways of realizing value from data at their fingertips. However, it is found that while collecting data is not difficult, value creation is often a big challenge. First, sensors and sensing techniques have been advancing rapidly for real time data collection with good enough accuracy. However, without efficient and effective ways to select, co-relate, integrate and transform multiple streaming data and their context information into manageable knowledge, these data are actually burdens instead of potentials to their owners. Second, despite numerous successful research efforts in data mining and machine learning, it is found that much less emphasis is put in the incorporation of domain knowledge into the data mining and pattern discovery processes, and in the use of behaviour genotypes such as loyalty and purchase power of customers to support final decision making. Third, related to the analytics platform, internet-of-things, service and cloud computing techniques are quite mature, and lots of machine learning algorithms are also widely available in both commercial and open source packages. However, how to use them for vertical service composition to provide “intelligence-as-a-service” for a given domain is still open for exploration. In this presentation, we will go into details of the current challenges of big data analytics and describe how behaviour analytics on trajectory data can help realize the value creation process from Big Data. In the discussion, both the science questions behind and the potential applications will be emphasized.
