An Efficient Sentimental Analysis Mining for Unclassified Data in Big Data Analytics

In the community networks, large numbers of users contribute to their opinions, for tracking and analyzing public response for making a valuable platform. Such tracking and analysis can provide critical information for decision-making in a variety of domains. In this paper, we generate a real-time suggestion for user comments among social network. To analyze the user comments, pattern recognition and data dictionary have been implemented. A raw dataset has been used as the input like posted by, date, time, posted topic, user name, and user comments. The data dictionary administrator can create a dataset for pattern matching, as it contains a group of all sentimental words. Sentimental words may be positive and negative words to be updated by the administrator. Instead of reading all the comments commented by the users, chart illustrates the graphical information of the discussion. Through the pattern recognition method, the user comments are categorized and the given input so that they could be separated into various sentiments like happy, sad, angry, etc. Moderate comments analyzed based on non-identified words from all the sentimental classifications. All the comments are grouped using global patterning reports and various charts generated. Ranking can be calculated using global patterning report. Through the generated chart, the admin can view a clarity report for the users’ opinion on the comment posted.