[Adapt] Our Last Seminar is 2-4 PM Next Wednesday in Room 414!

Kenny Zhu kzhu at cs.sjtu.edu.cn
Fri Dec 13 22:40:59 CST 2013

Dear ADAPTers,

Our last seminar will be next Wednesday. This last seminar will be special. I have invited two professors from University of Texas at Arlington and Hong Kong Polytechnic University to give two talks side by side. Because this seminar is open to the public, I have made a little change to the timing. Our seminar will begin at 2 PM and end roughly at 4 PM. The room is *414* instead of our usual 528. Please do make an effort to come and listen to the talks to support my visitors. 

As a tradition, we will take a group photo at the end of the seminar at 4 PM outside our lab at 341. In case you can't come to the talks, make sure you come to the photoshooting at 4 PM.

Below is the abstracts and bios of the two talks. Posters have been placed around the building and on the department website as well.

Have a good weekend!


Title: Parallel Analytics as a Service

Time: Wednesday, Dec 18, 2013, 2 PM
Venue: SEIEE-03-414
Host: Kenny Q. Zhu (朱其立)


Recently, massively parallel processing relational database systems (MPPDBs) like Vertica, Microsoft PDW, and Greenplum, have gained much momentum in the big data analytic market. With the advent of hosted cloud computing, we envision that the offering of MPPDB-as-a-Service (MPPDBaaS) will become attractive for companies having analytical tasks on only hundreds gigabytes to some ten terabytes of data because they can enjoy high-end parallel analytics at a cheap cost. In this talk, I will introduce Thrifty, a prototype implementation of MPPDB-as-a-service. The major challenge of building Thrifty is how to achieve a lower total cost of ownership by consolidating thousands of MPPDB tenants on to a shared hardware infrastructure while keeping certain performance SLA guarantees.  The talk will cover the system design, algorithms, and the latest updates of Thrifty.


Eric Lo is an associate professor of Hong Kong Polytechnic University, Department of Computing.  He started his PhD study at ETH Zurich in 2006 and obtained his PhD degree in 2007.  Before he returned to Hong Kong, he was an engineer of Google and a visiting scientist of Microsoft.  His main research interests are database system and cloud computing.  He has been PC members of all major top database conferences including SIGMOD2014, VLDB2014, and ICDE2014.  He has also served as reviewers of all major top database journals including TODS, VLDBJ, and TKDE.  His work has twice selected as bests of conferences (VLDB2005 and ICDE2012).

Title: Tackling Usability Challenges in Querying and Exploring Entity Graphs

Time: Wednesday, Dec 18, 2013, 3 PM
Venue: SEIEE-03-414
Host: Kenny Q. Zhu (朱其立)


We witness an unprecedented proliferation of entity graphs that capture entities (e.g., persons, products, organizations) and their relationships. Real-world entity graphs include knowledge bases, social graphs, citation graphs, drug and disease databases, and program analysis graphs, to name just a few. Users and developers are trying hard to tap into entity graphs for numerous applications, including search, recommendation systems, business intelligence and health informatics.  Both users and application developers are often overwhelmed by the daunting task of understanding and using entity graphs. The challenges lie in the gap between complex/big data and non-expert users. In retrieving data from entity graphs, the norm is often to use structured query languages such as SQL, SPARQL, and those alike. However, graph data is not “easier” than relational data in either query language or data model. If querying “simple” tables is difficult, aren’t complex graphs harder to query?

In this talk, I will introduce my group's ongoing efforts in tackling the usability challenges in querying and exploring entity graphs. Specifically, I will discuss GQBE, a system that queries graphs by examples and TableView, a technique that generates preview tables for entity graphs. I will also give an overview of our projects on computational journalism and entity query/exploration in Web text.


Dr. Chengkai Li is an Associate Professor in the Department of Computer Science and Engineering at the University of Texas at Arlington. He received his Ph.D. degree in Computer Science from the University of Illinois at Urbana-Champaign in 2007, and an M.E. and a B.S. degree in Computer Science from Nanjing University, in 2000 and 1997, respectively. After graduation in 2007, he became an Assistant Professor in the Computer Science and Engineering Department of UT Arlington and was promoted to Associate Professor in 2013. Dr. Li's research interests are in the areas of database, data mining and information retrieval, with the current emphasis on building large-scale human-assisting and human-assisted data and information systems with high usability, low cost and applications for social good. In particular, he works on computational journalism, crowdsourcing and human computation, database exploration by ranking (top-k), skyline and preference queries, database testing, entity search, query and exploration, query processing and optimization, usability challenges in using entity graphs, and Web data management. Dr. Li's papers have appeared in prestigious database, data mining and Web conferences including SIGMOD, VLDB, ICDE, EDBT, KDD, WWW, WSDM and CIKM, as well as in several leading journals such as TKDD and TKDE. He has served in the organizing committee of IEEE IPCCC several times (as General Co-Chair in 2012 and Program Co-Chair in 2010) and in the program committees of premier conferences such as VLDB, ICDE, EDBT, WWW, CIKM and ICDM. He has also been a reviewer for multiple prestigious journals, e.g., TODS, TOIS, TKDE and VLDB Journal. Dr. Li is a recipient of the 2011 and 2012 HP Labs Innovation Research Award.

More information about the Adapt mailing list