W&M Featured Events
This calendar presented by
William & Mary
[PAST EVENT] Aaron Koehl - Dissertation Defense - Computer Science
April 8, 2015
1pm - 2:45pm
Location
150
Abstract
Web services have continually improved since their inception in the early 1990's. Evolving from a simple protocol used to fetch linked static resources from a web server, contemporary web interactions rely upon dynamic capabilities in nearly every request. Database-backed, templated, programmatic execution technologies are collectively referred to as Web 2.0, and are employed by all of the top 50 heaviest websites on the Internet.
While static web technologies have reached a level of stability with regard to optimizations in server performance, dynamic web sites offer much greater degrees of both variability and complexity, and as a result, there are many competing opportunities for optimizing their complex interactions. We identify traditional web clients as active consumers of content originating from desktop machines, powered by commodity web browsers. In this dissertation, we explore scalability and performance implications of supporting non-traditional web clients. Non-traditional web clients include active content consumers from mobile devices, as well as passive content consumers such as search engines.
To address the burgeoning demand on servers placed by non-traditional clients, we present three middleware approaches. For active consumers, we address support for mobile devices, and present m.Site, a resource-efficient framework for content adaptation on mobile devices. For passive consumers, we study the behavior and impact of supporting search engine crawlers, identify optimizations for this class of request, and propose a middleware approach to mitigate what we term search engine overload. For both active and passive consumers, we study a technique for optimizing the image server with fidelity reduction using select psychovisual enhancements.
Bio
Aaron Koehl is a candidate for the Ph.D. in Computer Science at the College of William & Mary. Aaron received his Master of Engineering in Systems Engineering from the University of Virginia in Charlottesville, VA, and Bachelor of Science in Computer Science from Christopher Newport University in Newport News, VA. Aaron is a full-time Instructor of Computer Science and currently serves as Program Director of Information Systems and Information Science at Christopher Newport University.
Web services have continually improved since their inception in the early 1990's. Evolving from a simple protocol used to fetch linked static resources from a web server, contemporary web interactions rely upon dynamic capabilities in nearly every request. Database-backed, templated, programmatic execution technologies are collectively referred to as Web 2.0, and are employed by all of the top 50 heaviest websites on the Internet.
While static web technologies have reached a level of stability with regard to optimizations in server performance, dynamic web sites offer much greater degrees of both variability and complexity, and as a result, there are many competing opportunities for optimizing their complex interactions. We identify traditional web clients as active consumers of content originating from desktop machines, powered by commodity web browsers. In this dissertation, we explore scalability and performance implications of supporting non-traditional web clients. Non-traditional web clients include active content consumers from mobile devices, as well as passive content consumers such as search engines.
To address the burgeoning demand on servers placed by non-traditional clients, we present three middleware approaches. For active consumers, we address support for mobile devices, and present m.Site, a resource-efficient framework for content adaptation on mobile devices. For passive consumers, we study the behavior and impact of supporting search engine crawlers, identify optimizations for this class of request, and propose a middleware approach to mitigate what we term search engine overload. For both active and passive consumers, we study a technique for optimizing the image server with fidelity reduction using select psychovisual enhancements.
Bio
Aaron Koehl is a candidate for the Ph.D. in Computer Science at the College of William & Mary. Aaron received his Master of Engineering in Systems Engineering from the University of Virginia in Charlottesville, VA, and Bachelor of Science in Computer Science from Christopher Newport University in Newport News, VA. Aaron is a full-time Instructor of Computer Science and currently serves as Program Director of Information Systems and Information Science at Christopher Newport University.
Contact
[[vlthompsondopp]]