User-Customizable Health Pattern Detector Framework: Twitter Analysis Example

Lianna M. Hall, Kevin K. Nam, Jason Thornton, Marianne DeAngelus, Timothy J. Dasey

Abstract


A general-purpose method for automatic detection algorithm reengineering based upon Twitter keyword queries using user relevance/irrelevance feedback has been demonstrated to have superior performance and versatility compared to more static detection methods. A demonstration of the capability with an initial user interface has been performed. An extension of the processing that includes initial query term expansion prior to application of the customized detection is being investigated.


Full Text:

PDF


DOI: https://doi.org/10.5210/ojphi.v7i1.5798



Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org