Abstract:
Social sites are a platform where colossal bulk of data is shared, individuals as well as organizations share their information on various online media platforms. Among these sites Facebook is a valuable source where tremendous set of information is erected and can be evaluated for helpful outcomes. Facebook is the most utilized public network by wide-ranging population for information distribution. Facebook can likewise be very helpful for making people cognizant of healthcare informatics. This research is comprised of two parts, in first part two approaches were used to constitute different datasets for perusal of how Facebook pages of health organizations impact in distribution of healthcare information. A manual dataset was generated by accumulating Facebook posts of different health associations working in Punjab Pakistan. This data was statistically analyzed and based upon this data predictions were made for upcoming health relevant issues. An automated dataset was amassed by collecting health related posts of medical management associations. This information was further processed and statistically examined, also a classification tree was designed for the prediction of impending health problems. After the predictive analysis end results of both datasets were compared for accuracy, manual data set gave the accuracy rate of about 70% and automated dataset’s prediction was nearly 50% accurate. In second part of the research some factors were contemplated that made gigantic impact while accessing data from Facebook. The most important factor was the changing privacy policies of Facebook, as the cyber laws are modified the access to data is also restricted which results in less accuracy of the results. When security policies of Facebook are changed the access to maximum number of attributes is restricted and data mining techniques get effected. Limited access to Facebook data badly effects the data mining as well as the analysis process which results in low accuracy of the study. Few more factors that can effect this type of study were the selection of specific health organization, bulk of irrelevant information on health related pages and deficient users’ interaction to informative posts.
