首页 > 代码库 > 《textanalytics》课程简单总结(2):topic mining
《textanalytics》课程简单总结(2):topic mining
coursera上的公开课《https://www.coursera.org/course/textanalytics》系列,讲的很不错哦。
1、“term as topic”有非常多问题:
2、Improved Idea: Topic = Word Distribution:
3、定义问题(Probabilistic Topic Mining and Analysis):
4、解决这个问题之道(Generative Model for Probabilistic Topic Mining and Analysis):
– Model data generation with a prob. model: P(Data |Model, λ)
– Infer the most likely parameter values λ* given a particular data set: λ* = argmaxλ p(Data| Model, λ)
– Take λ* as the “knowledge” to be mined for the text mining problem
– Adjust the design of the model to discover different knowledge
当中:λ=({ theta1, …, thetak }, { π11, …, π1k }, …, { πN1, …, πNk })
5、The Simplest Language Model(generative model): Unigram LM
通过独立的生成每个词进而产生文档,因此:
? p(w1 w2 ... wn)=p(w1)p(w2)…p(wn)
? 參数为: {p(wi)} ,且 p(w1)+…+p(wN)=1 (N is voc. size)
? Text = sample drawn according to this word distribution,比如:
p(“today is Wed”) = p(“today”)p(“is”)p(“Wed”) = 0.0002 * 0.001 * 0.000015
6、两种预计文本产生概率的办法:
?最大似然预计
“最好”意味着“样本数据的似然值达到最大”:。
问题是,样本一般较小。
? 贝叶斯预计
“最好”意味着“和‘先验’一致,同一时候能非常好解释样本数据”,即Maximum a Posteriori (MAP) estimate。
问题是,怎样定义“先验”。
7、多个Unigram Language Model混合(以两个为例):
8、Probabilistic Topic Models: Expectation-Maximization (EM) Algorithm
样例:
9、Probabilistic Latent Semantic Analysis (PLSA)
本质思想:
数学关系:
PLSA中的EM:
11、LDA
内容參考:
http://blog.csdn.net/mmc2015/article/details/45009759
http://blog.csdn.net/mmc2015/article/details/45010307
http://blog.csdn.net/mmc2015/article/details/45011027
http://blog.csdn.net/mmc2015/article/details/45024447
《textanalytics》课程简单总结(2):topic mining