Matthew Kolb's Projects
iDiscuss: iDiscuss is a joint project with several professors at Mississippi State University to better understand student interactions within online discussion systems. More info: http://idiscuss.msstate.edu
Connection Diversity Measurement Tool: The tool was developed with Dr. Ed Allen and Dr. T.J. Jankun-Kelly at Mississippi State University to determine if information theory based metrics and visualization apply to network intrusion detection by analyzing network traffic logs. It requires developing a Java and C application for the Army Research Laboratory's (ARL).
MADAAT: The Microarray Data Annotation and Analysis Tool was a year long effort to aid the MSU Forestry Department with genetic research on poplar trees by providing a data storage and reporting mechanism for gene expression data in C#. The team consisted of 9-13 team members at any given time.
Intelligent Music Processing, April 2009: Project and paper aimed at grouping songs into a set of moods or genres, rating songs quickly, and provide ratings for new songs based on overall musical tastes. Methods focused on a graph of the songs with similarity metrics between songs and using graph partitioning for selecting genres and nearest neighbor methods for ranking songs.
An Analysis of Detailed Software Design Documentation Methods for Object-Oriented Code, April 2009: A summary of the importance of various software documentation qualities and an analysis of current tools capabilities in meeting these qualities is done by examining previous works and modern documentation methods.
Graph-Based Music Discovery, December 2008: The paper describes research based on implementing a graph-based system for visualizing a small world network of artists from a user's music collection, artist meta-data, and associated recommendations from Last.fm.
A Comparison of Programming Languages, December 2008: This term paper explains some of the key differences between the Ruby and Scheme programming languages. An emphasis on the readability, writability, and reliability will be used for evaluations.