IRAN Datasets

We conduct experiments with the datasets of SemEval 2014 task4 to evaluate our model, the SemEval 2014 datasets consist of reviews in two categories: Restaurant and Laptop, and the reviews contains three labels of sentiment polarity: {positive, negative, neutral}我们采用SemEval 2014任务4（SemEval 2014 Task4）的数据集开展模型评估实验。SemEval 2014数据集包含两类评论，分别隶属于餐厅领域与笔记本电脑领域，且该数据集的评论涵盖三种情感极性标签：{积极、消极、中性}