NC Health Info and Go Local: An Analysis of Web Change Impacts on Metadata Quality and A Proposed Framework for Semi-Automatic Metadata Maintenance
This paper reports on a two-part study focusing on automatic metadata maintenance for NC Health Info, a health information web portal for North Carolina residents. The first examination is an analysis of web changes and their potential impacts on metadata quality. The second examination proposes a framework for semi-automatic metadata maintenance, which builds off of a hierarchy of evaluation proposed by Greenberg (2005). This work is supported by a baseline counting of the average number of Dublin Core Metadata elements on Dublin Core website and a conceptual framework enhanced with mathematical Set Theory. The results lead to a model that enables catalogers to review only a portion of the web resources with detected changes. The proposed approach aims to benefit the metadata generation and maintenance by reducing human effort and time consumption while ensuring high metadata quality.
School:University of North Carolina at Chapel Hill
School Location:USA - North Carolina
Source Type:Master's Thesis
Keywords:web change detection consumer health website subject metadata maintenance quality evaluation
Date of Publication:04/07/2008