Hello, my name is Tony Snodgrass. I'm a graduate student at the University of Georgia pursuing a Masters degree in Artificial Intelligence (MSAI) and a Masters of Linguistics specializing in Language Variations (discourse and corpus analysis). My research interests include Natural Language Processing, Knowledge Representation, Robotics, and Neuroscience.

Formerly, I was a Lab Technician for the Institute for AI, supervised by Dr. Michael Covington. The Lab Tech position involved maintaining the general health and well-being of the computers in the IAI. Motherboards, monitors, soundcards, harddrives, projectors, servers, backups, Windows, F-Secure, group policy, active directory, user accounts, passwords, profiles, permissions, plants, plagues, people... blah blah blah. Basically, computers broke and I repaired them. My current position is as the SysAdmin/Programmer for the Linguistic Atlas Project supervised by Dr. Bill Kretzschmar.

I did my undergraduate studies at the University of Florida double majoring in Computer Science and Philosophy. While I was at UF, I racked up a lot of course work in Psychology, but not enough for a third degree. My Philosophy major was the best decision I've ever made.

My interests include watching a lot of movies, reading more books, compulsively sewing patches onto my backpack, and riding my motorcycle. Unfortunately, I was hit by a truck in April of 2007, and my knee was damaged. My favorite movie of all time is Ghostbusters. The books that I enjoy more than any others are the volumes of Calvin and Hobbes (created by a philosophy major, Bill Watterson). When I'm really bored, I do jigsaw puzzles. I like a few anime series, like Full Metal Alchemist, Rurouni Kenshin, Trigun, Cowboy Bebop, and Samurai Champloo. However, most of my time is spent on my school work, research, and a few other projects listed below.

My most recent projects have been for my Intro To Robotics course under Dr. Don Potter. Our class is divided up into four teams who work as groups to build robots for desired tasks. My teammates are Eric Drucker and Darren Earl, two very talented computer science undergrads. We've been way too engulfed in our projects to pick a team name. So far, we've concentrated mainly on purely reactive autonomous robots that perform simple tasks. The tasks that we perform are determined by the "challenge" put forth by Dr. Potter at the beginning of each project. Our first two challenges restricted us to the use of the Lego Mindstorms sets, v2.0 and earlier. Now, for the third challenge, we've been allowed to work with the Evolution Robotics ER1 robotics platforms, which is basically a laptop computer serving as the brain for a robot frame the laptop is mounted on. Click the picture to the left or click here to get to a summary page of our work. We hope to have demo videos up for all our robots soon.

Now that our robotics class has come to a close, I'm continuing to play with robots in the lab. My intention is to do a final project in Machine Learning that uses neural networks in series to interpret sensor data from 24 ultrasonic sensors for the purposes of producing a 8' X 8' grid map of my robot's surroundings. Provided that I get some good results, this should tangent well into a thesis on SLAM (simultaneous localization and mapping) using this technique.

Sylvia Plath was a brilliant poet and author who suffered from severe depression, leading to her tragic suicide on February 11, 1963 at the age of 31. She is most well-known for a book of poems Colossus and her novel The Bell Jar. In recent years, The Unabridged Journals of Sylvia Plath were published.

I'm currently working as an unpaid researcher for the CASPR project (Computer Analysis of Speech for Psychological Research) under Dr. Micheal Covington, and I've taken on my first research project. I'm going to be doing a computational linguistic analysis of the journals she kept during the last 12 years of her life. Her biography is well-researched documenting the full extent of her psychiatric treatment, including electroshock therapy. I cannot go into full detail on this project as it will most likely burgeon more than several different papers. Any and all publications from this project will be listed here.
Update: The publishing house of The Unabridged Journals of Syvlia Plath has been a little difficult to work with. This project has been postponed until I can scan and OCR the text from my copy of the book.

In 2004, my senior design project for my Computer Science degree was a program that solves Rubiks Cubes. This program is named Dr. Ray's Cube Solver after Dr. Greg Ray, who kindly agreed to be my project advisor (after I offered to name the program after him). The finished project was written in prolog and implemented with a crude java applet GUI that invoked the prolog engine via a jpl interface. The program averaged, out of 250 test trials, 43 moves for a solution with an upper bound of 57 moves and an execution time of under two minutes. My work on this project has not stopped since graduating. Since then, I've had many new ideas on how to both optimize the solutions as well as speed up execution time. This project involves complex search methods as well as the challenge of representing the state of the cube in a way that can be acted upon by the computer based on the rules a Rubiks Cube is bound by in real-life. I believe that I can get the execution time under one minute and average solutions down to approximately 34-36 moves. Eventually, I hope to design a better GUI for the front-end and publish Dr. Ray's Cube Solver online for free use. I plan to work on this project extensively in my free time.
Update: We've just put up a new server designated LEIBNITZ on our system which I, in my capacity as Lab Technician, plan to configure to run large prolog applications and scripts. If I can integrate a nice PHP user interface, then I think it'll work out nicely.

A friend of mine, Daniel Tuohy, did a fascinating project: utilizing genetic algorithms, his program can take a piece of music (written to be played on certain instruments) and rework it so that the same music can be played on instruments that the composer never intended the music to be played on. An example might be playing Mozart on an electric guitar (that's just an example I made up). Tuohy has written several papers on this project and presented and various international conferences. The only hiccup for Tuohy is that he must punch in the notes one-by-one read from sheet music, which sometimes takes hours. He told me that he's always wanted to write a java program that would read the sheet music for him, but he's never had the time to do it himself. Lucky for him, I needed a good idea for a final project in my Computer Vision class. So my contribution to Tuohy's work is going to be just that: by the end of Spring Semester 2006, I will give Tuohy a java function that takes a grayscale bitmap image of variable size as input and returns a string of notes in a format of Tuohy's specification as output. I think I know how the object recognition is going to work, but Tuohy is going to have to teach me how to read sheet music.
Update: Finished the project. It is written in both Java and C++. However, neither of the implementations get an accuracy higher than 94% on any test music. Noise is too unpredictable in scanned images. So, unfortunately, it is pretty-much useless to Tuohy. On a high-note (no pun intended) I got an A.

For my Final Project in my Natural Language Processing course as well as tandum to my work with the CASPR research group, I'm assisting Dr. Covington in implementing a more refined version of Cati Brown's reputed Computational Idea Density Rater. The original was implemented in the Python scripting language and in Prolog. The Python portion of it is affectionately known as "The Monty Tagger" using the Penn Treebank tag set. Dr. Covington used the work done by Turner and Greene in 1974 to refine Cati's original idea, and I have implemented this new counted tag set for him. The new density rater is done in Java and is more complete than the first one in hopes of providing a better count of propositions/ideas in a given text according to Turner and Greene's scheme. A paper is pending with Cati Brown as first author, me as second, and Dr. Covington as third. Dr. Covington has named my part of the project "The Full Monty Tagger".
Update: The Idea Density Rater, aka CPIDR has been developed near completion and tested on Turner and Greene's sample sentences. An abstract has been submitted to the LSA for a poster submission for their annual conference to be held in Anaheim, CA with Cati Brown as first author and me as second.
Update: v2.1 of CPIDR, a .net app written in C# was presented by myself at the Georgia Graduate Student Inderdisciplinary conference March 25th. The development of a distributable, installable version is very handy.
Update: v3.0 of CPIDR, another C# instantiation, has been written by Dr. Covington and put under a UGA copyright. Development is still an on-going process.

My Personal Links:

• Tony's Quotes - A page of my favorite quotations.
• Tony's Laws - A page containing a list of "laws" I've accumulated.
• Tony's Cartoons - A page full of funny cartoons and comics.
• Stages of Exam Anxiety - I found this too hilarious to not keep it.

• IMBIA - I attend weekly meetings of this student organization.
• UF Dept of Philosophy - Link to my old stomping ground.
• ComingSoon.net - A great resource for movie news and release information.
• The Facebook - Facebook me!
• Thinkgeek.com - A great place for gadgets and Tshirts.
• Newegg.com - One-stop shopping for all your computer component needs.
• MapQuest - The only people I trust for directions.
• Galileo - The BEST resource for academic journals and articles.

I was born Anthony Christopher Snodgrass to Micheal and Bobbie Snodgrass on December 24th, 1981 at Hahn AFB, Germany outside of Frankfurt. Our family is in the Air Force, so we moved around a lot.

My mother is Dr. Bobbie Snodgrass, PhD in Education, and my father pinned on as Brig. General Micheal Snodgrass in February of 2005. I have two sisters, my younger sister Kimberly Snodgrass (born in Texas) and my older sister Angela Ornelas who is married with two children. We had to move around a lot, so we're very adaptive people. With our last name, we have to be. My parents and my little sister currently live in Germany.

In my first highschool in Valdosta, GA I was a bit reclusive. I was an officer on our highly successful debate team, was a member of the Math and Science clubs, played soccer, and did one drama production of the play Greater Tuna. In my second highschool in Fairfax, VA I played football and tutored 7th grade math. I had several jobs in highschool: New World Bagelry and Cafe with Planet Smoothie, Waldenbooks, and Baskin Robbins. I worked the first job until their new liquer license required them to fire all minors, and I worked at Waldenbooks at the Valdosta Mall up until we moved to Fairfax, VA. In Virginia, I worked scooping ice cream at Baskin Robbins Springfield for 15 months until I left for college.

I did my undergraduate years at the University of Flordia in Gainesville. I dual majored in Computer Science and Philosophy. I attempted to pick up a third major in Psychology and desired a minor in Linguistics, but there is a cap to the number of credits a non-honor student is allowed to take. While at UF, I was Vice President of the UF Undergrad Philosophy Society as well as a member in good standing of Campus NOW, the UF Chapter of NOW. I completed my studies in the Fall of 2004.

While in Gainesville, I worked in several different places doing very different jobs. I did not work my freshman year. My first job was as waiter for a BBQ restaurant called Red Hot & Blue the summer I was back in Fairfax visiting my family. Sophomore year I was hired as a cook/bouncer at a bar'n'grill called Calico Jacks Seafood and Oyster Bar. I cooked, washed dishes, and dealt with fake IDs and other such unpleasantness on Quarter-Beer Tuesdays. When the long hours and low pay started to affect my GPA, I quit and replaced that job with two others. My first job was as a Desk Assistant for the UF Department of Housing watching a dorm front desk at night on weekends. The second job was a day job at an Equine Farm working with horses, but mainly just cleaning up and fixing fence boards. I worked there for about six months. Afterwards, I worked the Housing job fulltime, about 40 hours a week while I finished up my undergraduate work.

When my time at UF was finished, I got a job at LifeSouth Community Blood Centers working in the IT Department developing software. It was my first paid software job. Yep, I worked in a blood bank. Vampire jokes were not in short supply amongst my friends. It was a great job. I learned a lot of new skills and met a lot of great people. I encourage everyone to donate blood. And if you want to be a man about it, just walk in and ask for a bucket and a gun.

Having left Gainesville, I'm happily settled into Athens at my new school. The AI Department already feels like home the same way the UF Philosophy Department once did. School will most likely keep me here through the Spring of 2008, as I'm not trying to rush myself through my studies like I'm running a marathon. I'd been targetting the AI program since 2002, so I'm not going to waste my chance. My plan is to finish up and pursue additional graduate studies elsewhere. If I can, I want to do one of those study-intensive 10 month graduate programs overseas. Afterwards, on to my PhD. I don't know where this will be, but I'm hopeful for a good school. I got very very lucky getting into UGA, so I have to work extra-hard as I doubt lightning will strike twice for me.