Date of Award

2012

Degree Type

Thesis

Department

Electrical and Computer Engineering

First Advisor

Sridhar,Nigamanth

Subject Headings

Podcasts, Podcasting, Electronic data processing -- Quality control, podcast, podcasting, data quality, TDQM

Abstract

Podcasting has emerged as an asynchronous delay-tolerant method for the distribution of multimedia files through a network. Although podcasting has become a popular Internet application, users encounter frequent information quality problems in podcasting systems. To better understand the severity of these quality problems, we have applied the Total Data Quality Management methodology to podcasting. Through the application of this methodology we have quantified the data quality problems inherent within podcasting metadata, and performed an analysis that maps specific metadata defects to failures in popular commercial podcasting platforms. Furthermore, we extracted the Really Simple Syndication (RSS) feeds from the iTunes catalog for the purpose of performing the most comprehensive measurement of podcasting metadata to date. From these findings we attempted to improve the quality of podcasting data through the creation of a metadata validation tool - PodCop. PodCop extends existing RSS validation tools and encapsulates validation rules specific to the context of podcasting. We believe PodCop is the first attempt at improving the overall health of the podcasting ecosystem

Download
COinS