【摘要】 面对海量数据带来的冲击,传统的单机版贝叶斯分类程序存在处理的数据集有限、内存瓶颈和耗时较长等问题.本文通过对朴素贝叶斯模型进行研究,设计并实现了一种基于Map Reduce的朴素贝叶斯文本分类算法.实验表明,该算法具有较好的扩展性和加速比,可适用于海量密集文本分类.更多还原

【Abstract】 In the impact of the massive data, problems such as limited dataset, limited memory and time-consuming appeared in stand- alone Bayesian classification architecture. By researching of Naive Bayesian model,we design and realize an algorithm based on Map Reduce Naive Bayesian model. Some experiment results indicate that this algorithm can be expanded and accelerated, suitable for massive text classification.更多还原