Prediction of aspect category on customer reviews using association rule mining

Now a days online consumer review is a most powerful tool for decision making. This term serves as electronic word of mouth which become increasingly popular. Millions of people are now buying products and services via online. Web services are provided this feature to users openly. The web can provide an extensive source of consumer reviews. The user can read all the reviews and evaluate fair view of product or service. This can apply only to a limited number of reviews presented on the web. The web contain more than hundreds of reviews then problem arrived and time consuming also. A text processing framework is desirable which summarize all the reviews. This framework would find out general aspect category addressed in all review sentences. The method presented in this framework which applies association rule mining on co-occurrence frequency data to find out these aspect categories. From this result, generate polarity score for each aspect category. This polarity score helps to evaluate fair decision making for the customer as well as the company. The graph representation is also provided by the system for quickly evaluate the decision for products or services provided by the web.