Produced Data of Naive Bayes Sentiment Classifier

This is the data produced by the running of the Naive Bayes classifier algorithm. It is a list of every word in the vocabulary of the classifier, as well as the number of occurrences of each word, as well as the likelihood ratio of this word. Please note the likelihood ratio is calculated by taking the likelihood of word given a positive label divided by the likelihood of a word given a negative label. This data is licensed under the CC BY 4.0 international license, and may be taken and used freely with credit given. This data was produced by two different datasets, using a Naive Bayes classifier. These datasets were the Polarity Review v2.0 dataset from Cornell, and the Large Movie Review Dataset from Stanford.

本数据集为朴素贝叶斯分类器（Naive Bayes classifier）算法运行所生成的数据。其内容涵盖该分类器词表中的全部词汇、各词汇的出现频次，以及对应词汇的似然比。需特别说明：此处似然比的计算方式为，给定正标签时该词汇的似然除以给定负标签时该词汇的似然。本数据集采用CC BY 4.0国际许可协议（CC BY 4.0 international license）授权，标注原出处后可自由取用与使用。本数据集基于两套不同的原始数据集，通过朴素贝叶斯分类器生成：其一为康奈尔大学的Polarity Review v2.0数据集，其二为斯坦福大学的Large Movie Review Dataset大型电影评论数据集。