Dependency Parser to Analyze Customer Reviews on Products to identify Users Opinion

摘要

In today's world of dot com or internet where window shopping is being replaced by online shopping, it is very easy to find customer review of any product or about any services on internet Customer can make use of this information to take wise and more profitable decision while purchasing ? product or availing a service or making an investment. Similarly Opinions are very much essential for any organization's development and success. Opinion affects the quality of various products in great way. It'll indicate how much the customer has satisfied with the product along with its strength and weakness. There are a number of ways of collecting opinions; it may be through a feedback register, or through forums and other online feedback survey facilities. Because of this rich use of web communication, we find large amount of customer reviews on a product, due to this it becomes difficult for the customer and as well as for the manufacturer to analyze this huge data to find positive and negative opinion about a product. This needs a system which should automatically analyze these reviews and it should classify them as positive and negative. So we need a proper automated tool that can actually read through the comments . arid classify them according to the needs. The tool should be capable of going through a large set of comments in a short time and should be efficient enough to recognize almost any sentence we can throw at it. An attempt has been made to build this tool using Stanford Dependency Parser.