Data Assistant

Xiaoyue Elaine Wang

PhD Candidate Campus Address

Data Mining Lab Department of CSE

*****@**.***.*** Univ. of California Riverside

www.cs.ucr.edu/~xwang/ Riverside, CA 92521

1-951-***-**** United States

OBJECTIVE

Software engineering position with special interest in the area of data mining, machine learning

EDUCATION

09/2006- present Ph.D. (Expected 01/2010)

University of California, Riverside, USA

Major: Computer Science, Advisor: Prof. Eamonn Keogh, GPA: 3.9/4.0

09/2002- 06/2006 B.S.

Wuhan University, Wuhan, China

Major: Computer Science, GPA: 89.1/100

RESEARCH INTERESTS

Data mining/Machine learning techniques with applications to time series data

Image retrieving and analysis

Shape indexing/matching problems

Social network analysis

PROFESSIONAL EXPERIENCE

Sprint Applied Technology Lab Burlingame,CA 06/2008 - 09/2008

Research Intern

Social network analysis

Considered the problem of assigning weights to edges in a social graph such that the edges accurately reflect

the relationship between nodes. Compared various edge weighting techniques, ranging from simple aggregate-

volume metrics to time-series analysis of node interaction data for the node label prediction problem.

University of California Riverside Riverside CA 07/2007 - present

Research Assistant

Similarity Measures for Time Series Data

-Introduced an advanced version of Dynamic Time Warping algorithm which can improve the classification

accuracy when the time series data contains a certain amount of noises on both ends.

-Produced an augmented Euclidean distance measure for time series of shapes.

Multi-feature Image Matching

-Proposed an efficient method to combine different similarity measures such as shape and color for image

matching.

Indexing Techniques with Constraint on Space and Time

-Introduced a just-in-time indexing technique to speed up the joins in massive real valued datasets, showed

applications in domains as diverse as sensor mining, blog indexing and historical manuscript mining.

-Implemented an anyspace indexing algorithm with applications to data mining.

-Showed a framework working in the scenario of in memory data, e.g. sensor data, based on the time

series bitmaps technique. Efficient classifiers in this case can be updated in constant time and space in the

face of very high data arrival rates.

Semi-supervised Learning Algorithms on Shape Matching

-Produced a more accurate semi-supervised learning algorithm which leveraged off a novel observation

about the effects of shape complexity on distance measures. Showed an effective way to determine the

complexity of a shape in the time series representation.

Demos

-Built a demo website on content-based image indexing algorithms for a large database of scientific

nematode images. It allows users to retrieve different information from the database, e.g. to find most similar

species for the given query nematode image. The website is built and maintained using HTML, Postgresql,

PHP and JavaScript.

-Created a GUI tool in Matlab that converts photos to the appropriate representation. Now the tool is used to

index petroglyph images for Department of Anthropology in UCR.

University of California Riverside Riverside CA 09/2006 - 06/2007

Teaching Assistant

Assisted students with computer problems, troubleshooted hardware and software problems.

SKILLS

Programming: C/C++, BASH scripting, PHP, HTML, Postgresql, Latex

Software: Matlab

Operating System: Mac OS X, Linux, Windows NT/2000/XP

AWARDS

Dean's Distinguished Fellowship Award, University of California, Riverside, 2006-2008

Teng Fei Special Scholarship Award, Wuhan University, China, 2005-2006

JOURNAL

Wang, X., Ye, L., Keogh, E and C. Shelton. Annotating Historical Archives of Images. (International Journal of

Digital Library Systems, to appear January 2010)

SELECTED PUBLICATIONS

Wang, X., Ye, L., Keogh, E and C. Shelton. Annotating Historical Archives of Images. (runner up best student

paper award JCDL 2008)

Ding, H., Trajcevski, G., Scheuermann, P., Wang, X. and Keogh, E. Querying and Mining of Time Series Data:

Experimental Comparison of Representations and Distance Measures. (VLDB 2008).

Ye, L., Wang, X., Yankov, D., and Keogh, E. The Asymmetric Approximate Anytime Join: A New Primitive with

Applications to Data Mining (SDM 2008)

Ye, L.,Wang, X., and Keogh, E. Autocannibalistic and Anyspace Indexing Algorithms with Applications to Sensor

Data Mining. (SDM 2009)

Kasetty, S., Stafford, C., Walker, P., Wang, X. and Keogh, E Real-Time Classification of Streaming Sensor Data.

(ICTAI 2008).

Zhu, Q., Wang, X., Keogh. E, Lee, S. and Rampley, T. Towards Indexing and Data Mining all the Worlds Rock

Art. In the 37th Annual International Conference on Computer Applications and Quantitative Methods in

Archeology (CAA 2009)

Wang, X. and Keogh, E. Finding Centuries-Old Hyperlinks with a Novel Semi-Supervised Learning Technique.

(JCDL 2009)

Zhu, Q., Wang, X. and Keogh, E. Augmenting the generalized hough transform to enable the mining of

petroglyphs. (KDD 2009)

Contact this candidate