Scalable Keyword-based Content Filtering Solution on Key-Value Platforms
关键字内容过滤问题在Key/value平台上的可扩展性方案研究
讲座时间:2012年9月24日 14:00
讲座地点:嘉定校区济事楼417
主讲人: 饶卫雄 博士
芬兰赫尔辛基大学博士后研究员
Post Doctor Researcher, the University of Helsinki, Finland
【Profile】
Weixiong Rao received the BS and MS degrees from North (Beijing) JiaoTong University, and Shanghai JiaoTong University, respectively, and received the PhD degree from The Chinese University of Hong Kong in 2009. He has years of industry working experience and proficient in enterprise computing platforms. Currently he is working as a Post Doctor researcher in the University of Helsinki, Finland. His research interests include large scale networked data and content systems, and energy-aware mobile computing.
【Abstract】
The Web 2.0 era is characterized by the emergence of a very large amount of live content. A real time and fine grained content filtering approach can precisely keep users up to date the information that they are interested. Using keywords is the most frequent approach for users to define their personal interests. Then, the key of the approach is to offer a scalable keyword-based match algorithm.
In this talk, I will report the recent results deployed on two different scenarios of key/value-based platforms: structured P2P networks and clusters of commodity machines. First, we respectively study (i) an optimization problem to minimize the overall content forwarding network traffic on the P2P networks and (ii) another optimization problem to maximize the content matching throughput on clusters of commodity machines. After proving that both optimizations are dual problems and NP-hard, we propose approximation algorithms and develop practical prototypes. The experimental results on real data sets show that the proposed algorithms greatly outperform the state of arts.