Mining e-commerce satisfaction sentiment through a bilingual model

Recently years have witness that e-commerce has become internationally phenomenon in which sellers will handle multi-language buyers. The international customers often tend to make purchase decisions based on online review comments and recommended products. Therefore, understanding the sentiment of online comments will help sellers and help to make customized recommendations with higher likelihood of user satisfaction. We develop bilingual tools that handle both English and Chinese user comments. Therefore, the user's satisfaction we obtained is more generic and more objective. This paper addressed several key issues: Chinese segmentation, data mining models and systematic design. Before we conduct experiment, we validated our implementation with well accepted data. The experiments showed effectiveness and efficiencies of our implementation.