In the era of set retrieval in the libraries, information retrieval evaluation was all about precision and recall based on relevant documents. In the 1990s, large-scale ranked retrieval was evaluated using recall-precision curves and average precision. The beginning of the 21st century saw the widespread use of more ranked retrieval metrics such as normalised Discounted Cumulative Gain (nDCG), which has been used not only in the research community but also for tuning commercial web search engines. However, there is more to evaluating search: this talk discusses some latest developments. Search is no longer about returning a single ranked list: it may provide direct answers or aggregated results to users; it may interact with the user; it may adopt different strategies depending on the output device (e.g. a mobile phone). Search is no longer about returning relevant documents: it is about returning relevant, nonredundant and timely information. Evaluation methods need to be designed accordingly.

Tetsuya Sakai received a Master's degree from Waseda University in 1993 and joined the Toshiba Corporate R&D Center in the same year. He received a Ph.D from Waseda University in 2000 for his work on information retrieval and filtering systems. From 2000 to 2001, he was a visiting researcher at the University of Cambridge Computer Laboratory. In 2007, he became Director of the Natural Language Processing Laboratory at NewsWatch, Inc. In 2009, he joined Microsoft Research Asia. He is a Programme Co-chair of NTCIR. He is on the editorial board of Information Processing and Management and that of Information Retrieval the Journal. His past responsibilities include: chair of the Asia Information Retrieval Societies Steering Committee; general co-chair of AIRS 2009; programme co-chair of SIGIR 2013 and OAIR 2013; chair of IPSJ SIG-IFAT; co-editor-in-chief of IPSJ TOD; and co-organiser of the 1CLICK, INTENT, ACLIA and CQA tasks at NTCIR. He has received several awards in Japan, mostly from IPSJ. Recently, he won the AIRS 2012 best paper award.

User’s behavior during search, such as querying, examining and clicking, has strong implications for the user’s information need. User behavior analyzing and modeling help search engines in many aspects, including spam detection, query understanding, ranking and evaluation, etc. In this talk, I will share some recent research progress on user behavior understanding and modeling to improve search engine performance and to monitor social life. We look into both large scale click-through data collected from a popular commercial search engine and laboratory based eye-tracking data which is commonly used in psychological research. Some interesting biases have been observed, such as vertical result attraction bias, users’ examining sequence bias, users’ credibility bias, etc. Furthermore, I will show some cases that how user’s searching behavior analyses can help us make better understanding to our physical world such as epidemic and financial trends prediction, and social events detection.

Dr. Min Zhang is an associate professor in the Department of Computer Science & Technology (DCST), Tsinghua University. She received her Bachelor and PhD degrees from DCST at Tsinghua University in 1999 and 2003, respectively. During the past years, she has visited DFKI Germany, City University of HongKong, Kyoto University, and MSRA as visiting researcher. Dr. Zhang specializes in information retrieval, Web user behavior analysis and machine learning. She has published more than 100 papers on important international journals and conferences, such as JASIST, JIR, SIGIR, WWW, WSDM, CIKM, etc. She has participated in TREC (Text REtrieval Conference) benchmarks as the team leader since 2002. Her team has continuously achieved multiple top performances during 10 years. She also contributed in INTENT tasks in NTCIR evaluation as task co-organizer from 2011 to 2013. Dr. Zhang serves as area chairs or senior PC members at CIKM and AIRS, and PC members at SIGIR, WWW, WSDM, KDD, ACL, etc. Currently she is also the executive director of Tsinghua University-Microsoft Research Asia Joint Research Lab on Media and Search, and the vice director of Tsinghua-Sohu Joint Research Lab of Search Technology.

Cloud computing, which has been considered as the long dreamed vision of computing as a utility, enables convenient, on-demand network access to a shared pool of configurable computing resources and provides X-as-a-service in a flexible on-demand manner. It motivates both individuals and enterprises to remotely store their data in the cloud thus relieves them from the burden of data maintenance and local hardware and software management. Although cloud computing makes data services more appealing than ever, as a consequence of the shift, data outsourced to the "honest-but-curious" cloud server are no longer under the data owner's direct control, which may put highly sensitive data at higher risk. To mitigate the security concerns of outsourced data, a straight-forward and effective solution is to encrypt them locally before outsourcing. However, data encryption brings new challenges to effective data utilization. In this talk, we will take a close look at these challenging issues in the state-of-the-art and address them carefully for the success of cloud service deployment. The first challenge is how to directly enforce secure access control and support selective dissemination via selective encryption, while minimizing key management cost and achieving scalability in access control policy updates. The second challenge is how to achieve efficient query over encrypted data via encrypted index, while guaranteeing query privacy and supporting efficient dynamic data updates. Finally, we discuss how to efficiently audit the correctness of the outsourced data and verify the correctness of query results, while protecting data privacy from unauthorized entities.

Zhiyong Peng received B.Sc from Wuhan University, M.Eng. from Changsha Institute of Technology of China in 1985 and 1988, respectively, and Ph.D degree from Kyoto University of Japan in 1995. He is a professor and vice dean of computer school, Wuhan University of China. He worked as a researcher in Advanced Software Technology & Mechatronics Research Institute of Kyoto from 1995 to 1997 and a member of technical staff in Hewlett-Packard Laboratories Japan from 1997 to 2000. His research interests include complex data management, web data management, trusted data management. He is a member of IEEE Computer Society and ACM SIGMOD, a member of Database Society of Chinese Computer Federation, vice director of VLDB School China. He is general co-chair of MINES2009,WAIM2011, DASFAA2013 and PC Co-chair of DASFAA2012, WISE2006 and CIT2004.