文本分词，记词频，做词云

a <- segmentCN("1.txt",returnType="tm")

b <- read.table("1.segment.txt",fill=T)

?table 

lecture=read.csv("1.segment.txt",sep=",",header=TRUE,fileEncoding="UTF-8")   

# 查看前几行，看是否有字符编码问题  

head(lecture);  

# 获取数据集长度  

n=length(lecture[,1]);  

print(n)  

# == 文本预处理  

res=lecture[lecture!=" "];  

#剔除URL  

res=gsub(pattern="http:[a-zA-Z\\/\\.0-9]+","",res);   

#剔除特殊词  

res=gsub(pattern="[我|你|的|了|是]","",res);       

# == 分词+频数统计  

words=unlist(lapply(X=res, FUN=segmentCN));  

word=lapply(X=words, FUN=strsplit, " ");  

v=table(unlist(word));    

# 降序排序  

v=rev(sort(v));   

d=data.frame(word=names(v), freq=v);   

# 过滤掉1个字和词频小于20的记录  

d=subset(d, nchar(as.character(d$word))>1 & d$freq>=20)  

#设置中文输出

par(family='STKaiti') 

mycolors <- brewer.pal(8,"Dark2")

wordcloud(d$word,d$freq,random.order=F,random.color=F,colors=mycolors)