Role of Online Data from Search Engine and Social Media in Healthcare Informatics

Introduction

Internet is now affecting and facilitating nearly every aspect of modern life, from healthcare and education to government and business. In past few years, healthcare organizations and professionals are using social media in order to promote, support and spread health related information and data for improving both personal and community health practices (Househ, 2013; Chretien & Kind, 2013). Moreover, the younger generation also uses Internet especially social media for research and making health related decisions. In a survey (Health Fact Sheet, 2015), Internet users of approximately 72% checked online for information on health in past year. Online health seekers of about 77% used search engine. 11% said that they looked for health information at specialized health information site such as WebMD. 3% started their health related research at sites like Wikipedia and an additional 2% said that they used social networking sites such as Facebook and Twitter.

For health related applications and monitoring, the idea of using online data came from the estimation of Influenza incidence using logs of health related search engine queries (Eysenbach, 2006). Studies in (Ginsberg et al., 2009; Hulth et al., 2009; Palet et al., 2009; Achrekar et al, 2011, Aramaki et al., 2011; Polgreen et al., 2008) show that epidemics trend can be detected with information available on Web and there is a strong correlation in the frequency of online search queries and tweets with epidemics events (Xu et al., 2011). Hence, the behavior or pattern of when and how Internet users search may provide early indications or clues related to future concerns and expectations. For example, analysis conducted by Ettredge et al., (2005) on jobs and jobs opportunities related keywords searched by users over Internet has generated an accurate and useful statistics on the unemployment rate. These studies also suggested that people suffering from any kind of disease or health issues uses World Wide Web (WWW) to search for disease information. Logs of search queries (or terms) entered in search engines can provide valuable information on health related issues, especially the detection and monitoring of emerging epidemic diseases, as it is possible to track changes in the volumes of specific search queries.

However search query data is noisy, coarse and it does not provide any contextual information. People search for information related to health on search engine for various reasons, such as concern about oneself, friends or families. Some searches are done because of general interest that is usually initiated by a live event, news report or new scientific discovery (Signorini et al., 2011). Furthermore, errors discovered in Google Flu Trends serves as reminder that this big online data paradigm required further critical investigation and the development of more empirical methodologies for exploring the predictive utility of Internet data (Lazer et al., 2014a; Lazer et al., 2014b). One other limitation is that researchers and scientists do not have full access to search engines logs. Recently, social media data have been used effectively for disease surveillance as they contain contextual health information with diverse descriptions of health states. A study on “Twitter stream” (Twitter Study, 2009) revealed that despite high level of noise, a major proportion of Twitter message contain informative, links to the useful information and news content or spam and self-promotion.