Applying a bilingual model to mine e-commerce satisfaction sentiment
Gongjun Yan, Wu He, Hui Shi, Danda B. Rawat · Journal of Management Analytics · 2014
In recent years we have witnessed that e-commerce has become an international phenomenon in which sellers often need to do business with buyers who use different languages. As customers, including international customers, often make purchase decisions based on online reviews, comments and recommendations made by other customers, understanding the sentiment of online comments becomes necessary for sellers who want to provide customized recommendations to users for a higher level of customer satisfaction. To this end, we have recently developed a bilingual model to handle both English and Chinese user comments posted on e-commerce websites. This paper addresses several key issues including Chinese segmentation, data mining models and systematic design. An experiment in mining user satisfaction sentiment with English and Chinese online user comments illustrates the value of our developed bilingual model.