Mining Comparative Relations of Products/Services from Web Opinions

This project proposes an approach to mine comparative relations from the web using text mining techniques. Numerous opinions exist on the web today, such as those which can be found in forums, blogs, and online shopping websites. These opinions usually have high values for businesses. Many efforts are being put to mine opinions from the web to support the enterprise management decisions and provide customers shopping assistances. A very important but rarely being explored aspect of this field is mining comparative relations between products/services. This kind of comparative relations are more useful for individual customers and enterprise managers. For example, before making a decision on purchases, customers always tend to refer to the opinions of comparison between two or more products from past buyers; moreover, managers always track the comparative opinions to evaluate their products’ merits and shortcomings in comparison with their competitors’ similar products.