The beginning of my studies in computer science coincides with rise in popularity of social networking sites. At that time, Facebook and MySpace were rapidly gaining in popularity, and I, along with many of my peers, was a user. Motivated by the rising popularity of social networks and a deep interest in the mathematical foundations of these services, I decided to study these networks. While reviewing the related literature, I was intrigued by the assertion that many real-world datasets contain interrelated entities and capture relationships among these entities. I further hypothesized that a link mining approach could be applied to social networks, protein interactions, etc., to predict the existence of links where none were annotated. Working on research projects in this area for the past three years, and contributing original ideas to several publications, has helped clarify my desire to conduct research in computer science.

Looking forward, I intend to study the dynamic, temporal and graphical nature of the link structure in various domains in order to predict probabilities of link existence. I am specifically interested in feature selection, extraction and construction within this domain. My perception is that while learning algorithms are well studied and understood, the discovery and extraction of features that drive these learning algorithms is not. Therefore, I intend to study the identification and extraction of features paying specific attention to those graphical features most relevant to link mining tasks.

Information Integration and Analysis Center

As a research associate at Kansas State University, I was the technical project leader and pricipal architect for a federally funded information extraction and synthesis project. Papers are forthcoming.

Multimodal Information Access and Synthesis

The Department of Homeland Security sponsored the Multimodal Information Access and Synthesis (MIAS) summer research institute. During the Summer of 2008 I attended and lead a research group in the creation a vertical search engine. This type of search engine was able to create a very specific model of Sports teams the University of Illinois Urbana-Champaign (UIUC). From this model we were able to provide novel search capabilities.

Text and Information Extraction

During research in Information Extraction I was getting poor performance because of unwanted text and ads were cluttering up my corpora. So I created a method of extracting content from web pages by looking at the organization of the underlying HTML code. This method was very successful; two papers were published describing the approach.

Link Mining in Social Networks

For my undergraduate research project, I assisted a team in link mining research performed on LiveJournal. This research resulted in several papers and a couple of theses.

Medical Information Retrieval

Worked on speech-based medical information retrieval software. Much of this information is proprietary, however publications are pending

Unstructured Decision Making

I teamed up with Dr. Jan Crow to develop scientific research studies that looked at how people make unstructured decisions. I wrote several applications that tested the decision making ability of survey participants and assistend Jan in interpereting the results. One paper was presented at the annual meeting of the Society of Judgment & Decision Making, and there are more to come.

Senior Design Project

WaveletVision

Artificial Intelligence Project

TAC Agent A Search-based Approach

Distributed Systems Project

Hood Sensor Abstraction

NSF Fellowship Essays

Research Experience [PDF]

Personal Statement [PDF]

Research Proposal [PDF]

NDSEG Fellowship Application

Application [PDF]