Enhancing trust in dental care recommendation systems : using trust-enhanced information from social networks to improve the matching between patients and dentists

Publication Type:
Thesis
Issue Date:
2016
Full metadata record
Files in This Item:
Filename Description Size
01front.pdf320.7 kB
Adobe PDF
02whole.pdf3.58 MB
Adobe PDF
The recent growth in social media has impacted the way users are searching and sharing health information online. Crowdsources, such as review and rating websites, provide an outlet for consumers to share their opinions on healthcare professionals. Yet, faced with the enormity and diversity of information across multiple online sources, finding the right information can be a challenge for users, particularly when there is no consistency in the evaluation criteria across various sources. This difficulty is manifested when existing review and rating websites do not take patient information into consideration. Extremely biased views – positive or negative – are capable of skewing recommendations and thereby compounding the situation. This makes it important to filter trustworthy information from health social networks and dental crowdsources. In the case of dental care, the invasive nature of many dental treatments highlights the importance of selecting a suitable trustworthy provider for many patients, who may be anxious or reluctant to visit a new dentist. By analysing, from multiple perspectives, the trustworthiness of information available to patients, this study proposes a new trust-enhanced information model for dental care recommendation systems. In this model, dentists are profiled based on subjective information extracted from dental crowdsources. Subjective qualities are also used to profile patients. Currently, online social network data cannot be used for profiling purposes due to privacy and identification concerns. Instead, one of the popular personality tests, the DISC personality test, is used in this study. The importance and suitability of subjective qualities for recommendations is explored. Two matching algorithms are evaluated based on the responses to an online survey. When the patients are classified based on their levels of fear, preferred search methods and other attributes, their list of recommended dentists changes. The subjective characteristics of both patients and dentists are important factors which need to be incorporated to improve the matching capability of dental care recommendation systems. Including the subjective qualities of users could change the way that recommendations are provided in the future, especially in the health sector where the wrong information can lead to adverse impacts on health. Although patients’ discussions about their health are sensitive and private, they can benefit from more accurate recommendations in relation to health care providers.
Please use this identifier to cite or link to this item: