Sarah K Tyler

I am either a Math Geek who likes words, or a Word Nerd who likes math.

For the past 4 years I've been working at Lawrence Livermore National Labratory in the area of Information Retrieval and Natural Language Processing: primarily in Entity Extraction, Document Clustering and Classification. I enjoy writing algorithms that search large corpora for a specific detail that satisfies the information need. Although many of the techniques I've employed have been knowledge-poor, I've done some research on different ways to model prior knowledge. I presented my research on Knowledge Inference over graphical models, to the Graph Analysis working group in April, 2006.

In the coming year, I'm hoping to branch into Personalized Search, as well as Hidden Web Information Retrieval and Federated Search. I am excited to start Graduate Studies in the fall at the University of California, Santa Cruz for my doctorate.

In my spare time I often write scripts and applications for my primary website, SKTyler.com. Most notably, I wrote a Writing Sample Analyzer, which predicts reading ease based on Flesch, Fog and Flesch-Kincaid approximations. The Analyzer receives approximately one hundred hits a day, and was donated to the Logan Utah Public Library for their Adult Literacy Group. I'm also do volunteer website design and development for other organizations when I have time.

Before my interest in Text, I interned at BodyMedia with their informatics group. I implemented machine learning techniques, such as cross-validation, and built models to detect different activities based on measurements recorded from an armband. Prior to that I interned at QVC.