以此分析台大的研究趨勢
#USPTO之搜尋網頁
url =
"http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=%2Fnetahtml%2FPTO%2Fsearch-adv.htm&r=0&p=1&f=S&l=50&Query=AN%2F%22NATIONAL+TAIWAN+UNIVERSITY%22&d=PTXT"
wordpage = read_html(url)
#連接進各專利說明書內並擷取其中的摘要部分作分析
xpath ="//table//tr/td[2]/a"
half_url=xml_attr(xml_find_all(wordpage,xpath),"href")
url_content = paste("http://patft.uspto.gov", half_url,sep="")
page1=" "
for( i in 1:50){
patent = read_html( url_content[i])
target = "//body//p[1]"
part = xml_find_all(patent,target)%>%xml_text()
page1 = paste(part,page1)
}
#文本清理
docs = Corpus(VectorSource(page1))
# Convert the text to lower cas
docs <- tm_map(docs, content_transformer(tolower))
# Remove english common stopwords
docs <- tm_map(docs, removeWords, stopwords("english"))
# Remove punctuations
docs <- tm_map(docs, removePunctuation)
# Eliminate extra white spaces
docs <- tm_map(docs, stripWhitespace)
# 因為法律文獻的書寫方式有固定模式，因此要大量移出不重要的名詞
docs = tm_map(docs,removeWords,c("generate","object","association","connected"," period","includes","shell","provided","first","second","present","end","group","formed","element","comprising","can","the"))
##產生詞頻表格與文字雲
dtm <- TermDocumentMatrix(docs)
m <- as.matrix(dtm)
v <- sort(rowSums(m),decreasing=TRUE)
d <- data.frame(word = names(v),freq=v)
print(head(d, 10))
## word freq
## layer layer 52
## metal metal 34
## method method 33
## electrode electrode 31
## invention invention 24
## gate gate 21
## region region 21
## structure structure 20
## substrate substrate 19
## catalyst catalyst 18
print(wordcloud(words = d$word, freq = d$freq, min.freq = 3,
max.words=20, random.order=FALSE, rot.per=0.38,
colors=brewer.pal(7, "Dark2")))
## NULL