Quantitative Measures of Online Health Information (QMOHI): Broadening the Impact through Improved Usability, Applicability, and Effectiveness
By: Miho Shimizu
Department: Computer Science
Faculty Advisor: Dr. Anagha Kulkarni
Universities and colleges offer medical and health services through Student Health Centers (SHCs), and SHC websites are a primary healthcare information source for university and college students. While educational institutions are trusted as reputed information sources by students, there is no guideline or standard for quality control of online healthcare information on SHC websites. Our tool, Quantitative Measures of Online Health Information (QMOHI), was designed to address the issue and provide data-driven quality assessments of various aspects of health information posted on SHC websites. QMOHI's assessments are based on eight different metrics, for a health topic of interest specified by the user. The user specifies the health topic of interest by providing a set of words that are representative of the topic (the keywords). QMOHI uses these keywords to identify web pages on SHC websites that may be relevant to the health topic and proceeds to download a copy of these web pages. The content of these web pages is then analyzed by the following eight quality metrics – Readability, Prevalence, Coverage, Objectivity, Polarity, Navigation, Timeliness, and Relevance. On top of this work, we added new features. The sentence highlighting feature was designed to help public researchers examine assessment results for their studies. Previously overlooked texts that appear in PDF contents or images have been included for evaluation by the latest upgrade. In addition, the estimation algorithm of the Relevance metric, which represents how similar the information on the SHC website is to a comparison document containing salient information about the health topic, has been revised and achieved improved accuracy. Finally, we added a powerful function that helps users prepare the keywords of a health topic of interest required to run QMOHI by suggesting keywords for the health topic. With these additions, QMOHI has enhanced its usability, applicability, and effectiveness.