引用本文: | 谭文堂,王桢文,殷风景,等.一种面向涌现的比较性话题模型[J].国防科技大学学报,2013,35(4):146-155.[点击复制] |
TAN Wentang,WANG Zhenwen,YIN Fengjing,et al.A comparative topic model for words burstiness[J].Journal of National University of Defense Technology,2013,35(4):146-155[点击复制] |
|
|
|
本文已被:浏览 8445次 下载 6229次 |
一种面向涌现的比较性话题模型 |
谭文堂, 王桢文, 殷风景, 葛斌, 肖卫东 |
(国防科技大学 信息系统工程重点实验室,湖南 长沙 410073)
|
摘要: |
提出一种CDCMLDA生成模型来实现跨文本集的话题分析,采用狄利克雷组合多项式模型(Dirichlet Compound Multinomial ,DCM)对文本集中词的涌现现象进行建模,把DCM模型和LDA结合起来分析文本集之间话题的差异,采用蒙特卡罗期望最大化方法进行参数推导。在多个实际数据集中通过定性和定量的方法对模型进行评价,实验表明,模型不仅能够发现不同文本集间的异同,而且在模型困惑度指标上相对当前两种主要跨文本集的话题模型具有明显的优势。 |
关键词: 比较性文本挖掘 涌现 话题模型 CDCMLDA模型 |
DOI: |
投稿日期:2012-12-18 |
基金项目:国家自然科学基金资助项目(60903225);湖南省自然科学基金项目(11JJ5044);国防科技大学优秀研究生创新基金项目(S100502) |
|
A comparative topic model for words burstiness |
TAN Wentang, WANG Zhenwen, YIN Fengjing, GE Bin, XIAO Weidong |
(Science and Technology on Information Systems Engineering Laboratory, National University of Defense Technology, Changsha 410073,China)
|
Abstract: |
State-of-the-art cross collections topic models suffer from the serious flaw that it cannot capture the tendency of words to appear in bursts. Based on LDA (Latent Dirichlet Allocation),a topic model CDCMLDA(Cross-collection Dirichlet compound multinomial Latent Dirichlet Allocation), which models the burstiness phenomena of words using Dirichlet compound multinomial (DCM) distribution, was proposed. A Monte Carlo Expectation Maximization algorithm for model inference was presented. A variety of qualitative and quantitative evaluations of CDCMLDA were performed, which shows that CDCMLDA not only discovers the common and unique aspects on topics, but also improves the model perplexity compared with the two cross-collection topic models. |
Keywords: comparative text mining burstiness topic model CDCMLDA model |
|
|
|
|
|