Hierarchical Agglomerative Clustering approach for Automated Attribute Classification of the Health Care Domain from User Generated Reviews on Web 2.0

摘要

Due to Web2.0, large masses of healthcare product opinions are freely available. These enormous amounts of opinion contain valuable information for industries, consumers, and upgrading healthcare of patients. One of the major challenges in healthcare domain to find out valuable information hidden in a large volume of data. Thus data science succors us as many kinds as a decision it quicker or better as well as. Thinking so most of the popular companies invest much money in data scientists so that they could get the right information to provide an accurate result. It is the hollow which can show us all detail as by data science. In this paper, we use Data Science Hierarchical Agglomerative Clustering (HAC) system for classifying adopting the contraceptive method. In this system, we are using Frequent Pattern-growth for association rule mining for frequent feature extraction and corpus, Part Of Speech (POS), and interaction information used for infrequent feature identification and opinion words extraction, therefore, it is required to use both features as well as sentiment word for extracting the orientation. We use Thesaurus and Negation rule to identify the polarity of sentiment words. Finally, Hierarchical agglomerative clustering is used to categorize the contraceptive method.