yassiracharki/Amazon_Reviews_Binary_for_Sentiment_Analysis

Amazon Reviews for Binary Sentiment Analysis数据集是从亚马逊评论中构建的，用于二元情感分析。评论分数1和2被视为负面，4和5被视为正面，分数3的评论被忽略。数据集中，类别1代表负面，类别2代表正面。每个类别包含1,800,000个训练样本和200,000个测试样本。数据以CSV格式存储，包含类别索引、评论标题和评论文本三列。评论标题和文本使用双引号进行转义，内部的双引号通过两个双引号转义，换行符通过反斜杠加n字符转义。

The Amazon Reviews for Binary Sentiment Analysis dataset is constructed from Amazon reviews for binary sentiment analysis. Review scores 1 and 2 are considered negative, and scores 4 and 5 are considered positive. Reviews with a score of 3 are ignored. In the dataset, class 1 represents negative and class 2 represents positive. Each class contains 1,800,000 training samples and 200,000 testing samples. The data is stored in CSV format with three columns: class index, review title, and review text. The review title and text are escaped using double quotes, internal double quotes are escaped by two double quotes, and new lines are escaped by a backslash followed by an n character.