Li Liu US Permanent Resident CONTACT INFORMATION  Email liungc gmail com San Jose California US by rogerholland

VIEWS: 141 PAGES: 5

									Li Liu
US Permanent Resident

CONTACT INFORMATION 
                                                                            Email: liungc@gmail.com
San Jose, California US                                                     http://www.utdallas.edu/~liliu


RESEARCH INTERESTS
Data mining, privacy preserving data mining, anomaly detection, information security, security metrics, and
security issues in data and data applications

EDUCATION  

            Ph.D. in Computer Science
                                                                                          May 2008
            University of Texas at Dallas, Richardson, TX
                 Dissertation Title: Perturbation Based Privacy Preserving Data
                   Mining Techniques For Real-World Data
                 Advisor: Dr. Bhavani Thuraisingham
                   Co-advisor: Dr. Murat Kantarcioglu

            M.S. in Computer Science Software Engineering,
                                                                                          December 2003
            University of Texas at Dallas, Richardson, TX

            B.S. Degree in Computer Science,
            Northern Jiaotong University, Beijing, China
     
    WORK EXPERIENCE  
                                                                                        Jan 2008 - Present
    EBAY INC., SAN JOSE, CALIFORNIA, US

    Information Security Engineer of Global Information Security (GIS)
    Department
         Lead Security Metrics engineer of Metrics & Reporting Team
         Identifying and development the successful security measurement across
          eBay corporation, site and adjacencies.
         Implementations of the automatic data feed to Security Metrics on Archer
          Smart Framework.
         Establishing and development plans to drive and track the remediation of
          vulnerabilities in major categories; monitoring and reporting the status of
          compliance of SOX, PCI and security policies.
         Development and coordination of overall Information Security process and
          Strategy.
         Providing the accurate metrics and benchmarks to target the security
          awareness training.
         Assistance the risk management including analysis, monitoring and
          migration.




1                                                                                                        Li Liu 
    SECuR-IT Program at EBAY INC.,
                                                                                                     Jun 2007 - Aug 2007
    Summer Research Seminars and Internship
    This is a very special program organized by Team for Research in ubiquitous
    Secure Technology (TRUST).
         Summer Experience, Colloquium and Research in Information
             Technology at Stanford University and San Jose State University (SECuR-
             IT)
         SECuR-IT is a graduate academic immersion with internship program
         Worked closely with eBay InfoSec Team to identify the Security Metrics
     
    CYBER AND DATA APPLICATIONS SECURITY LAB, UT DALLAS
                                                                                                     Jan 2005 - Dec 2007
    Teaching Assistant                                                                               Sep 2006 - May 2007
       Assist with courses including Data Mining, Machine Learning, Data and
        Application Security, and Trustworthy in Semantic Webs
       Provide guidance and direction for graduate students projects and term papers
    Research Assistant
                                                                                                     Jan 2005 – Dec 2007
    PHD RESEARCH PROJECTS:
        Privacy Preserving Data Mining (PPDM)
          Design and implement in C, Matlab and Java, January 2005 to December 2007.
          Survey the technologies and approaches in PPDM
          Examine the perturbation model-based approaches intensively
          Propose a link-based PPDM framework
          Propose a two phase perturbation model for PPDM
          Propose an adaptable perturbation model of PPDM
          Conduct a great amount of experiments using different data sets
          Investigate the applicability of perturbation model-based PPDM on real world
              data sets
          Propose a novel privacy preserving decision tree algorithm
    After investigating the approaches in PPDM intensively, we have proposed a two phase
    perturbation model. Based on this model, we propose an adaptable perturbation model
    because privacy is a personal choice that one size will not fit in all the situations. Most
    perturbation based approaches have an inter step which reconstructs the original data set’s
    distribution. We conducted a huge amount of experiments using different real life data, and
    hence point out this inter step may not work well for the real world data sets. The main
    reason is distribution is a hard problem. The results of reconstructing the original data
    distribution will rely on the original data set’s distribution complexity. Furthermore we
    have proposed a new approach in perturbation based model, which approach instead of
    reconstructing the original data distribution, map the data mining algorithm with the
    correction of the bias. We apply this technique in decision tree classifier, and build a novel
    privacy preserving decision tree classifier, which classifies directly on the perturbed data
    set. Our experimental results have shown that our algorithm get very good results working
    with real world data sets.




2                                                                                                                  Li Liu 
        Anomaly Detection with Privacy Constraints in Medical Data
          Design and implement in C, Java and Matlab, January 2007 to May 2007.                     Jan 2007 - May 2007
          Investigate the currently technologies in anomaly detection
          Apply privacy constraints with anomaly detection technique
         Anomaly Detection and Fraud Detections are two very important issues to Credit
         Card and Medical Insurance companies. Disclosure of sensitive information can lead
         to or cause serious problems, e.g. identity theft. The goal of this research is to tackle
         this challenging problem with privacy protection. After investigating different
         approaches in anomaly detection, we will apply several PPDM techniques, e.g. k-
         anonymouszation, randomization and perturbation, to the anomaly detection domain.
         We will combine and modify the approaches in both domains to suit the privacy
         preserving anomaly detection situation. We also will build a test bed using Texas
         Healthy Medical Data to test and refine our approach. Current research is ongoing to
         find a robust solution that can be used in Privacy Preserving Anomaly Detection.
    MASTER RESEARCH PROJECTS:
                                                                                                     Jan 2003 – Aug 2004
        Real-time Classification of Multi-attribute Motion Data Using Support Vector
         Machines
          Design and implement in C, Matlab and Java, April 2004 to June 2004
          Apply data mining in motion detection for American Sign Language
          Investigate the approaches in time series
        Automatic Image Annotation and Retrieval using Weighted Feature Selection
          Design and implement in C, Matlab and Java, April 2004 to August 2004
          Apply different weighting technique in image annotation
          Investigate the different techniques in image retrieval
        Intrusion Detection System Using Data Mining Bayesian Classifier
          Design and implement in Java, January 2003 to August. 2003
          Investigate and compare different approaches in intrusion detection using
             network log data set

    OPTIMAL SOFTWARE TECHNOLOGY, Dallas TX
                                                                                              Apr 2004 – Dec 2004
    Software Engineer Developer

       Generate test and validation plans against the Software Requirement
        Specification (SRS)
       Generate test cases and run test cases in different developing phases
       Write bug reports, track bugs and confirm the fixed bugs

    VANDA GROUP, Hong Kong, China
                                                                                              Jun 1996 – Dec 2000
    BEIJING NEW GENERATION COMPUTER GRAPHICS COMPANY,
    Beijing, China
    Vice President

       Set up new branch in printing equipment industry, Beijing New
        Generation Co.
       Marketed publishing and printing equipment retail network.
       Imported and distributed publishing and printing equipment in China
        market

 




3                                                                                                                   Li Liu 
COMPUTER SKILLS 
Languages:               C, C++, Java, HTML, Basic, Visual Basic, Assembly, Pascal
Operating Systems:       Windows 9X/NT/XP, UNIX/Linux, MS-DOS
Database:                Oracle SQL
Networking:              UNIX Socket Programming, TCP/IP
Tools:                   Matlab, Microsoft Office, Rational Rose, Eclipse, CVS, MS Front page, Flash


PUBLICATIONS 

JOURNAL PUBLICATION: 

[1].   “The Applicability of the Perturbation Based Privacy Preserving Data Mining for Real-world
       Data”, extended work of [6], Li Liu, Murat Kantarcioglu, and Bhavani Thuraisingham, in Data and
       Knowledge Engineering (DKE), an Elsevier journal, volume 65, issue 1, April 2008.


CONFERENCE AND WORKSHOP PUBLICATION: 

[2].   “Security Metrics In Governance, Risk And Compliance”, Li Liu, accepted in the Fourth
       Workshop on Security Metrics (MetriCon 4.0), USENIX Security Symposium 18th, August 2009,
       Montreal Canada.

[3].   “Security Metrics In A Web-Based Enterprise Environment And Lessons Learned”, Li Liu and
       Caroline Wong in NSF Workshop on Data and Applications Security, February 2009, Arlington
       Virginia.

[4].   “Privacy Preserving Decision Tree Mining from Perturbed Data”, Li Liu, Murat Kantarcioglu,
       and Bhavani Thuraisingham, in Hawaii International Conference On System Science 42(HICSS
       42), January 2009, Big Island, Hawaii. Nominated For The Best Paper Of The "Decision
       Technologies And Service Sciences" Track.

[5].   “Global Information Security (GIS) Metrics Enterprise Plans and Lessons Learned”, Caroline
       Wong, Li Liu, and Dave Cullinane, in the Third Workshop on Security Metrics (MetriCon 3.0),
       USENIX Security Symposium 17th, July 2008, San Jose, California, US.

[6].   “The Applicability of the Perturbation Model-based Privacy Preserving Data Mining for Real-
       world Data”, Li Liu, Murat Kantarcioglu, and Bhavani Thuraisingham, in ICDM 2006 Workshop
       on Privacy and Security Aspects of Data Mining, December 2006, Hong Kong, China.

[7].   “A Novel Privacy Preserving Decision Tree Algorithm”, Li Liu, Murat Kantarcioglu, and Bhavani
       Thuraisingham, in Technical Report in Computer Science department at UTD, October 2006,
       Report No UTDCS-51-06.

[8].   “An Adaptable Perturbation Model of Privacy Preserving Data Mining”, Li Liu, Bhavani
       Thuraisingham, Murat Kantarcioglu and Latifur Khan, in ICDM 2005 Workshop on Privacy and
       Security Aspects of Data Mining, November 2005, Huston, TX USA.

[9].   “A Link-based Privacy Preserving Data Mining Framework”, Li Liu, Latifur Khan, Bhavani
       Thuraisingham and Chris Clifton, in SCISS 2005 The South Central Information Security
       Symposium, May 2005, Austin, TX, USA.




4                                                                                                Li Liu 
[10].    “Real-time Classification of Multi-attribute Motion Data Using Support Vector Machines
        (SVM)”, Chuanjun Li, Li Liu and et al., in MDM/KDD2004 Fifth International Workshop on
        Multimedia Data Mining "Mining Integrated Media and Complex Data", August 22 - 25, 2004,
        Seattle, WA, USA.

[11].   “Automatic Image Annotation and Retrieval using Subspace Clustering Algorithm” Lei Wang,
        Li Liu and L. Khan, in International Workshop on Multimedia and Database (MMDB) 100-108,
        November 2004.

[12].   “A New Cluster Based Intrusion Detection” Qing Chen, L. Khan, and Li Liu et al., Technical
        Report in Computer Science department at UTD, July 2003, Report No UTDCS-32-03.

SELECTED SPEECH AND CONFERENCE PRESENTATIONS: 

[1].    Invited speech: “Security Metrics, Governance and Compliance Management”, Li Liu, Archer
        National Summit, won the Archer Innovation Awards, April 2009, Orlando, Florida.

[2].    “Privacy Preserving Decision Tree Mining from Perturbed Data”, Li Liu, Murat Kantarcioglu,
        and Bhavani Thuraisingham, in Hawaii International Conference On System Science 42(HICSS
        42), January 2009, Big Island, Hawaii. Nominated For The Best Paper Of The "Decision
        Technologies And Service Sciences" Track.

[3].    “The Applicability of the Perturbation Model-based Privacy Preserving Data Mining for Real-
        world Data”, Li Liu, Murat Kantarcioglu, and Bhavani Thuraisingham, in ICDM 2006 Workshop
        on Privacy and Security Aspects of Data Mining, December 2006, Hong Kong, China.

[4].    “An Adaptable Perturbation Model of Privacy Preserving Data Mining”, Li Liu, Bhavani
        Thuraisingham, Murat Kantarcioglu and Latifur Khan, in ICDM 2005 Workshop on Privacy and
        Security Aspects of Data Mining, November 2005, Huston, TX USA.

[5].    “A Link-based Privacy Preserving Data Mining Framework”, Li Liu, Latifur Khan, Bhavani
        Thuraisingham and Chris Clifton, in SCISS 2005 The South Central Information Security
        Symposium, May 2005, Austin, TX, USA.

 
 
SELECTED REVIEWER SERVICE 
Reviewer:
    Invited Reviewer of IEEE Transactions on Knowledge and Data Engineering, TKDE.
    Invited Reviewer of Journal of Management Information Systems.
    Invited Reviewer of Data and Knowledge Engineering (DKE), an Elsevier journal.

External Reviewer:
    External Reviewer, 33rd International Conference on Very Large Data Bases (VLDB), 2007.
    External Reviewer, the 2006 IEEE International Conference on Data Mining (ICDM), 2006.



REFERENCES 
   Upon request.




5                                                                                              Li Liu 

								
To top