基于背景学习的迭代文本分类框架.docx

基于背景学习的迭代式文本分类框架  PAGE \* MERGEFORMAT 40 题名（中英对照）： 基于背景学习的迭代式文本分类框架 Background Learning Based Iterative Framework for Text Classification 摘要 计算机及互联网的普及在带给人们的生活和工作丰富资源的同时也使信息呈现爆炸式增长。随着网络中文本数据呈指数级增长，信息的人工分类和管理也已经被计算机自动分类所替代。 经过多年的研究和发展，文本自动分类领域已经开发出一些相对成熟的算法。对于中文文本分类，也已经有一些方法和系统可以达到较高的分类准确率。研究分析发现，在文本预处理阶段，对于歧义语段的划分始终是影响分类准确率的一个重要因素，至今仍未完全解决。本文结合互信息度理论，提出一种基于背景学习的迭代式框架，以此为基础通过对分词数据预处理来改进传统的基于朴素贝叶斯模型的文本分类算法。 本文使用新浪网不同类别数据对提出的迭代式框架进行实验评估。实验结果表明本文提出的基于背景学习的迭代式文本分类框架可行有效。 关键词：背景知识，迭代，互信息度，朴素贝叶斯，文本分类，歧义消除 Abstract The popularity of computers and the Internet brings people rich resource for daily life and work. At the same time, the explosive information and data had over-flooded us. With the text in the Internet growing at an exponential rate, manual categorization has been out of date and replaced by automatic classification. A number of algorithms have been proposed after decades of research. For Chinese text classification, various approaches and systems have been developed, which can achieve relatively high classification accuracy. More specifically, in text pre-processing stage, the division of ambiguous phrase is a key factor that affects the accuracy. And it has not been properly solved. This paper presents a background-based iterative framework integrated with the mutual information theory. It is used into the data preprocessing to improve the traditional text classification algorithm which is based on the Na?ve Bayesian model. Data from various Sina categories are used for the experimental evaluation of the framework. The results show that the proposed background learning based iterative framework for text classification is feasible and effective. Keywords: Background Knowledge, Iteration, Mutual Information, Na?ve Bayesian, Text Categorization, Disambiguation 目录  TOC \o 1-3 \h \z \u  HYPERLINK \l _Toc389290315 摘要  PAGEREF _Toc389290315 \h I  HYPERLINK \l _Toc389290316 Abstract  PAGEREF _Toc389290316 \h II  HY