Ontology-based Sentiment Analysis Process for Social Media Content

ocial media provides a platform where users share an abundance of information on anything and everything. The information may consist of users’ emotions, feedbacks, reviews, and personal experiences. In this research a novel Ontology-based Sentiment Analysis Process for Social Media content (OSAPS) with negative sentiments is presented. The social media content is automatically extracted from the twitter messages. An ontology-based process is designed to retrieve and analyse the customers’ tweet with negative sentiments. This idea is demonstrated with the identification of customer dissatisfaction of the delivery service issues of the United States Postal Service, Royal Mail of United Kingdom, and Canada post. The tweets related to the delivery service include delay in delivery, lost package/s or improper customer services at the office in person or at call centres. A combination of technologies for twitter extraction, data cleaning, subjective analysis, ontology model building, and sentiment analysis are used. The results from this analysis could be used by the company to take corrective measures for the problems as well as to generate an automated online reply for the issues. A rule-based classifier could be used for generating the automated online replies