| 摘要 |
At present, many large-scale consumer goods or food suppliers would receive a lot of feedback data from consumers, which is vital to the improvement of corporate management. Nevertheless, these data fail to be fully utilized or labeled manually piece by piece due to employees’ poor skills or officers’ lack of awareness in most of these enterprises. This paper investigates a simple method of handling such data. First, several basic databases are built, including index system, industry topic lexicon and emotion lexicon. Then, the polarity judgment method is studied. Finally, criticism rate, appreciation rate and suggestion rate are studied and determined. All these rates have gained application in a large-scale enterprise of China and received good effect. |