Databases @ CMU
The databases group at Carnegie Mellon University focuses on high performance database
architectures, multimedia, and data mining. We participate in a number of cross-disciplinary
efforts, and closely collaborate with a number of other groups at CMU.
DB alumni mentioned in the MIT technology review!
18 Dec, 2013
Prof. Hanghang TONG (MLD phd) and Prof. Leman Akoglu (CSD phd) worked on analyzing the correlation between quality of questions and quality of answers, in ``stack-overflow'' - their paper is on the web: http://arxiv.org/abs/1311.6876
The MIT-tech-review article can be found at
Congratulations, Hanghang and Leman!
U Kang is chosen as finalist, for the KDD-dissertation award!
26 Jul, 2013
Prof. U Kang, SCS Alumni, is chosen as the finalist for the 2013 KDD Dissertation Award, which is the highest honor for a data mining thesis. His dissertation is titled "Mining Tera-Scale Graphs: Theory, Engineering and Discoveries" and it focused on the award winning PEGASUS system for finding patterns and anomalies in billion-node graphs using thousands of machines. U will receive a certificate of recognition during the opening ceremonies at the upcoming KDD Conference in Chicago.
SCS team, alumni and colleagues attract the 'best paper' award in CIKM'12!
1 Nov, 2012
The paper gives fast, scalable algorithms to strengthen (or weaken) a network, by careful edge additions (or removals), and gives applications on real graphs.
The full citation is:
Hanghang Tong, B. Aditya Prakash, Tina Eliassi-Rad, Michalis Faloutsos and Christos Faloutsos
Gelling, and Melting, Large Graphs by Edge Manipulation, Maui, Hawaii, USA, Oct. 2012.
url of pdf:
The first two authors are SCS alumni (2010 and 2012, resp.), advised by the last author, SCS faculty Prof. Christos Faloutsos.
CIKM is one of the top conferences in information retrieval and data mining with 1099 articles submitted this year, and only 146 of them accepted as full papers.
Polo wins dissertation distinction!
24 Oct, 2012
Prof. Polo Chau wins the SCS Honorable Mention for his dissertation. Polo
defended his dissertation with title "Data Mining Meets HCI: Making Sense
of Large Graphs" in MLD in August 2012. Warmest congratulations Polo!!
Lei Li wins distinction!
16 Jul, 2012
Dr. Lei Li (SCS, phd'11) won the prestigious KDD best dissertation award (runner up),
for his dissertation ''Fast Algorithms for Mining Co-evolving Time Series''.
KDD is the premier Data Mining conference, with high competition from the best research groups
in the world.
Lei's thesis developed novel algorithms for forecasting, clustering and missing-value imputation
for time sequences in a broad spectrum of settings, from motion-capture sequences
to data-center monitoring to computer network ``BGP'' traffic mining.
Warmest congratulations Lei!
Mr. Alex Beutel wins NSF Fellowship!
24 Mar, 2012
We are delighted to announce that SCS doctoral student Mr. Alex
Beutel was awarded the 2011 NSF Graduate Research Fellowship,
(to be applied 2012-2015). The amount is $30,000 per year for up
to three years.
Polo has an invited article in ACM Crossroads!
22 Apr, 2011
Polo speaks out on graph mining and fraud detection
Polo Chau has an invited article in ACM Crossroads,
explaining how his work can help spot fraudsters
U Kang wins a ICDM 2010 NSF Student Travel Award!
5 Nov, 2010
U Kang is selected for the NSF Student
Travel Award to attend the International Conference on Data Mining conference 2010 in Sydney, Australia. The award includes
US $865 to cover the expenses incurred for attending the conference.
PEGASUS graph mining system wins 'open source software' award!
13 Oct, 2010
SCS team wins 3rd position (silver prize) in the 'Open Source Software World Challenge 2010',
among 26 competing submissions. The team consists of
graduate SCS students Mr. U Kang and Mr. Polo Chau, and advisor Christos Faloutsos,
with several more contributors, listed on the project web site
The award includes $2000, plus travel expenses to accept the prize.
The PEGASUS system is able to mine billion-node graphs, using parallelism
and specifically, 'hadoop'. Code, documentation, instructions video and related papers
are on the web site.
Congratulations U and Polo!
Leman Akoglu wins a Google ECML/PKDD 2010 Conference Grant!
24 Aug, 2010
Leman Akoglu wins the Google Conference and Travel grant for the ECML/PKDD 2010 that is offered to two female students to encourage more female computer scientists to attend and participate in the ECML/PKDD conference.
The grant includes free registration for the main conference and up to 300 Euros for travel costs from Google.
Leman Akoglu and Mary McGlohon win the 'best paper' award at PAKDD 2010
13 Jul, 2010
SCS PhD students Leman Akoglu and Mary McGlohon receive the 'best paper' award at PAKDD 2010. The paper is titled OddBall: Spotting Anomalies in Weighted Graphs by Leman Akoglu, Mary McGlohon and Christos Faloutsos and gives fast algorithms to spot strange nodes in large social networks.
PAKDD is one of the top data mining conferences. The paper was selected among 412 submissions, and 42 accepted full-papers.
Congratulations, Leman and Mary!
U Kang and Babis Tsourakakis: ICDM'09 Best Application Paper Award Runner Up
13 Nov, 2009
SCS graduate students U Kang and Babis Tsourakakis attracted the 'best applications paper award' (runner up), in ICDM'09, one of the top data mining conferences, for their paper
PEGASUS: A Peta-Scale Graph Mining System - Implementation and Observations
by U Kang, Charalampos (Babis) Tsourakakis, and Christos Faloutsos
ICDM 2009, Miami FL.
The paper was selected among 70 accepted papers, out of a total of 786 submissions,
and it showed how to use 'hadoop' and Yahoo's M45 machine, to analyze one of the largest
publicly available graphs (over 100Gb).
Moreover, the paper has been invited for fast-track possible publication to the KAIS journal.
Congratulations, U and Babis!
Leman Akoglu wins the 'best paper' award in PKDD09
24 Sep, 2010
We are delighted to announce that Leman AKOGLU attracted the
Best Knowledge Discovery Paper award in PKDD 2009,
out of 105 papers that were accepted in the conference,
and out 422 submissions. The paper was also accepted for
publication in the Data Mining
and Knowledge Discovery journal, for the ECML PKDD 2009 special issue.
The paper is titled
RTG: A Recursive Realistic Graph Generator using Random Typing
by Leman Akoglu and Christos Faloutsos, ECML PKDD, Bled, Slovenia, Sep. 2009.
ECML/PKDD is one of the top data mining conferences.
The paper presents an intuitive and fast method of generating
realistic graphs. It also proves analytically or empirically that
the generated graphs match 11 properties of real graphs,
making it the most realistic generator so far.
Fan's paper among the best in this year's KDD!
8 Jun, 2009
"BBM: Deriving Click Models from Petabyte-scale Data"
in KDD'09, by Chao Liu, Fan Guo, Christos Faloutsos
is selected as one of the best in this year's KDD, and invited for fast-track
journal publication. KDD and ACM-TKDD are, respectively,
the premier conference and journal for data mining.
The paper analyzed unprecedented volumes of click-stream
data, and discovered methods to do better ranking of search results.
Fan, a PhD student in SCS, did this work during his summer internship at MSR
last year, under the guidance of Dr. Chao Liu.
Leman's paper among the best in ECML/PKDD'09
6 Jun, 2009
RTG: A Recursive Realistic Graph Generator using Random Typing
by Leman Akoglu and Christos Faloutsos
was accepted to ECML/PKDD'09, and, as one of the best 14 papers,
is invited for immediate journal publication, out of 422 submissions
and 105 acceptances.
Leman is a PhD student in SCS, and ECML/PKDD is one of the top
conferences in data mining. The paper describes a simple, intuitive
mechanism to generate realistic graphs, and it shows that the resulting
graphs obey numerous of the patterns observed in real graphs.
Jure wins the ACM SIGKDD Dissertation Award!
21 May, 2009
Prof. Jure Leskovec wins the prestigious ACM SIGKDD dissertation award!
Jure defended his dissertation in MLD in September of 2009.
The title is "Dynamics of Large Networks", describing numerous, surprising patterns
that hold for real graphs, as well as several powerful tools to study blogs, instant-message networks
social networks like LinkedIn, and other types of real graphs.
ACM SIGKDD is the premier society for data mining research,
and the above award is the highest distinction for a Ph.D. dissertation in the area.
The award, which constitutes a $2,500 honorarium and a plaque, will be presented during the opening ceremonies
at the upcoming SIGKDD Conference in Paris, Sunday, June 28. Jure will have 15 minutes during the ceremony
to present a summary of his work.
Polo wins Yahoo KSC Program Award
15 Apr, 2009
Mr. Polo Chau wins a Yahoo Key Technical Challenge grant!
The grant is awarded to a limited number of exceptional Ph.D. students
and it carries a gift of $5,000 unrestricted funds to support their research activities.
The award also includes an exclusive invitation to the KSC graduate student Summit,
where Polo will have the chance to get feedback and interact with some
of the best Yahoo scientists.
Polo wins 2009 Symantec Fellowship!
9 Feb, 2009
Mr. Duen Horng (Polo) CHAU wins a 2009 Symantec Fellowship, second year in a row!
Polo is in the MLD PhD program, working on visualization and anomaly detection for large graphs.
The Symantec Graduate Fellowship program was established to identify and support exceptional graduate students who are investigating novel ways to solve real-world problems in protecting security and availability. In addition to the financial support, Polo will have the opportunity to work directly with Symantec’s top researchers on new investigations and will also have access to security insight from Symantec’s global intelligence network via the option of a summer internship
Babis' paper among the best in ICDM'08
28 Jan, 2009
Charalampos (Babis) Tsourakakis
had a paper that was invited for fast-track publication, as one of the best
in ICDM'08. Babis is a PhD student at SCS/MLD.
ICDM is one of the top data mining conferences,
and the paper gives an extremely fast method (1000x faster) to compute triangles in social networks. The details of the paper are
"Fast Counting of Triangles in Large Real Networks, without counting: Algorithms and Laws"
by Charalampos Tsourakakis
ICDM '08, Pisa, Italy, Dec. 2008
Jimeng Sun, SCS alumni '07, wins the prestigious 'best paper' award in ICDM'08
7 Jan, 2009
ICDM is one of the top data mining conferences. The paper proposed MET (Memory Efficient Tucker) decomposition, to handle large sparse tensors, a vital problem for time-evolving graphs, among other applications. The paper citation is:
T. G. Kolda and J. Sun, "Scalable Tensor Decompositions for Multi-aspect Data Mining," in Proc. IEEE International Conference on Data Mining (ICDM), 2008.
Jimeng had already attracted the 'KDD best dissertation award', for his thesis on tensors and streams.
Jure Leskovec in the news!
30 Aug, 2008
SCS PhD candidate Jure Leskovec, and his MSR colleagues, studied the largest social network in published literature, the Microsoft Instant Messenger network, and found 6.6 degrees of separation, among other fascinating results. The work is mentioned in major news venues, including Washington Post, MSNBC, BBC news, Guardian, Spiegel.
The work showed that, even in a huge network of millions of people, we still have the ``six degrees of separation' phenomenon, that Milgram observed decades ago, in a social network of about 100 people.
Jimeng Sun wins runner-up for SIGKDD best dissertation award
1 Jul, 2008
Dr. Jimeng Sun (Ph.D. CMU-CSD 2007) attracted the runner-up award for the best SIGKDD dissertation. Jimeng's dissertation is on tensor and stream mining, proposing novel and efficient methods to handle streams of numerical data, as well as streams of graphs.
He applied his methods on chlorine monitoring in the drinking water (joint project with Prof. Jeanne VanBriesen of CIT/CMU), on monitoring the self-star data center of PDL/CMU (with Prof. Greg Ganger and his group), and also on monitoring computer traffic (with Prof. Hui Zhang, SCS/CMU)
Jimeng will receive a plaque and will be recognized during the award ceremonies at the upcoming ACM SIGKDD Conference in August 24-27.
Furthermore, he will present his work during the conference. The ACM SIGKDD conference is the most selective and prestigious data mining conference.
Warmest congratulations, Jimeng!
Polo Chau attracts publicity
23 Jun, 2008
Polo Chau attracts publicity in popular media. WJS-online, CNN-Money, Pittsburgh Tribute-Review and others, mention Polo and the prestigious fellowship he received from Symantec. Congratulations, Polo!
Hanghang Tong, Spiros Papadimitriou, Christos Faloutsos and colleagues win the SDM08 Best Paper Award
1 May, 2008
Hanghang Tong (CMU), Spiros Papadimitriou (CMU alumni), Christos Faloutsos (CMU) and colleagues win BEST PAPER AWARD at the SIAM Data Mining conference, 2008.
The paper was authored by Hanghang Tong (CMU), Spiros Papadimitriou (IBM and CMU Alumni), Philip Yu (IBM) and Christos Faloutsos (CMU) with title Proximity Tracking on Time-Evolving Bipartite Graphs. SIAM-DM is one of the top data mining conferences.
The work focuses on social networks, and specifically on measuring the proximity of nodes, as the networks change over time. With careful design, the proposed methods achieve up to 2 orders of magnitude faster computation over straightforward competitors.
Congratulations, Hanghang, Spiros and Christos!
Mary McGlohon wins a Yahoo Key Technical Challenge grant!
10 Feb, 2008
The Key Technical Challenges Program is a new Yahoo! program that provides a limited number of exceptional PhD students with $5,000 each of unrestricted funds for the support of their research activities". The award also includes an exclusive invitation to a Yahoo! organized workshop that will focus on novel disciplines and important technical challenges for the Internet research community.
Polo Chau wins Symantec Research Labs Fellowship
6 Feb, 2008
Polo Chau wins a Symantec Research Labs Fellowship, for 2008-2009. The fellowship covers tuition, stipend, a laptop preloaded with Symantec software, and an invitation to spend a summer internship at the Symantec Research Labs. This dovetails perfectly with the fraud detection work that Polo did last year and attracted media attention.
For more info on the fellowship:
About this website