Mining Social Network Data for Enhanced Consumer Perception Measurement

Abstract

The Support Vector Machine (SVM) is a prevalent machine learning method which is employed for classification, regression, and many statistical tasks, adept at producing a hyperplane or several hyperplanes in high-dimensional or infinite-dimensional environments. An SVM analyzes datasets and can be trained iteratively to identify patterns of similar behavior or shared features among events within the database. In this study, a new data mining tool was developed to evaluate patients with type 2 diabetes mellitus and their familiarity with Sitagliptin. To achieve this, a two-stage research framework was designed. The first stage involved self-organizing mapping (SOM) for exploratory research, which identified mechanisms based on user preferences expressed in blog entries. Consumer clusters have positive or negative drug views.  Network analysis revealed notable forum users in the second round. This modeling provided insights into influential users and their role in shaping drug-related discussions. The results of this study open new avenues for rapid data acquisition, user feedback, and analysis to improve public health outcomes. Furthermore, the insights offer valuable input for healthcare providers and pharmaceutical suppliers.