Influences of Serendipity on Consumer Medical Information Personalization

Public Deposited

Downloadable Content

Download PDF
Request Version for Screen Reader
Last Modified
  • March 19, 2019
Creator
  • Fan, Xiangyu
    • Affiliation: School of Information and Library Science
Abstract
  • Serendipity is an important concept in the field of information science. It has a potential of enhancing information seeking process by unexpected discovery. Serendipitous recommendation has been incorporated into the design of personalized systems to minimize blind spots in information delivery. Little evidence has been found to identify how serendipity influences personalization of consumer medical information delivery. This dissertation attempts to examine what roles serendipity plays in filtering consumer medical information and to understand how to incorporate serendipity in an effective manner. In addition, the study seeks to clarify user attitudes on unexpected discoveries of medical content in filtering settings as well as users' interest changes during this process. To empirically analyze the influence of serendipity, a medical news filtering system named MedSDFilter was developed. The system can personalize the delivery of news articles based on users' interest profiles. In MedSDFilter, serendipitous recommendation was integrated into personalized filtering through one of three serendipity models (randomness-based, knowledge-based and learning-based). Using Medical News Today site as information source, three different system modalities were compared by conducting user experiments. Thirty staff members were recruited to read and rate medical news delivered by one of three system modalities. The results of user study indicate that serendipity has an important role in medical news content delivery. As for how to incorporate serendipity, it is shown that using physician knowledge effectively enhanced serendipitous recommendation. In addition, the results suggest that the performance of serendipitous recommendation was further improved after learning algorithms were adopted. This study also provide some evidence to show user satisfaction on unexpected discovery and user interest change associated with this type of discovery. Finally, the study demonstrated the individual difference in seeking consumer medical information. The results of this study provide the system designers implications and suggestions to avoid potential drawbacks related to over-personalization in information delivery. This study enhances the understanding of users' behavior regarding the consumption of medical information and generates new guidelines which can be used in developing information systems in medical area.
Date of publication
Keyword
Subject
DOI
Identifier
Resource type
Rights statement
  • In Copyright
Advisor
  • Capra, Robert
  • Quiroga, Luz
  • Mostafa, Javed
  • Hemminger, Bradley M.
  • Greenberg, Jane
Degree
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2015
Language
Publisher
Place of publication
  • Chapel Hill, NC
Access right
  • There are no restrictions to this item.
Date uploaded
  • June 23, 2015

Relations

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details Fan_unc_0153D_15433.pdf 2019-04-11 Public Download