A Lexicon-Enhanced Method for Sentiment Classification: An Experiment on Online Product Reviews

Abstract

As an emerging communication platform, Web 2.0 has led the Internet to become increasingly user-centric. People are participating in and exchanging opinions through online community-based social media, such as discussion boards, Web forums, and blogs. Along with such trends, an increasing amount of user-generated content containing rich opinion and sentiment information has appeared on the Internet. Understanding such opinion and sentiment information has become increasingly important for both service and product providers and users because it plays an important role in influencing consumer purchasing decisions.