Aspect based Parsing for Sentiment Analysis in Big Data

Today online reviews become enormously valuable sources for mining customers' opinions on services and products. Extracting these opinions from these reviews and hulling the gold bar out of them are hard-won task. Though it is a herculean, it has great crunch on the decision making process of the companies and consumers. This is the reason for sentiment analysis to be a crowd-pleasing topic of research. There are many techniques proposed for improving the accuracy of the sentiment analysis using parts-of-speech (POS) approach. But the authors present an argument that POS tagging has time and space complexity in sentiment analysis and proposed a novel algorithm, Aro_Jen to enhance the performance of sentiment analysis without POS process. Aspect and sentiment based lexicons are used for experimental analysis to prove the claim of the time and space complexity of POS tagging process in sentiment analysis with twitter dataset. Result of this research has the potential of being successfully applied to eliminate POS-tagging process in other text classification problems.