ABSTRACT

Internet of Health Things (IoHT) has turned out to be a challenging application of the Internet of Things (IoT) and cloud computing. In IoHT, the computation and communication of healthcare data are carried out to monitor the health condition of the patients. The improvement in applying Internet and social networks has activated users to interchange views, sentiments, and thoughts. This exchange of data leads to a path for sentiment analysis (SA). The key objective of SA is classifying data into three stages: positive (P), negative (N), and neutral (NU). Here, an efficient MapReduce-based Hybrid Decision Tree Classifier-Term Frequency-Inverse Document Frequency called the MPHDT-T method is presented for mining user's sentiment. An MPHDT-T is applied for classifying information depending upon the polarities of every context in social network data. The polarity value is estimated with the application of emotion corpus as well as a diabetic corpus. It helps in analyzing the relationship of food habits, external events, as well as diabetes risk aspects in India by employing social networking data. More than 2 million data are explored for research and it is limited in India. The simulation outcome exhibits that the MPHDT-T process is effective in a multimode cluster. The simulation outcome shows that there is no single factor related to diabetes risks and a set of typical issues involved in diabetes.