R语言文本分析

Posted 2021-04-18 管理研究方法

如果对实验研究感兴趣，欢迎阅读：；如果对问卷研究感兴趣，欢迎阅读：、、；如果对多层线性回归模型感兴趣，欢迎阅读：、、；如果对元分析感兴趣，欢迎阅读：、。

---

R code来源于政府报告词云图^[1]

分析的网络文本来源于政府工作报告（文字实录）^[2]

R语言分析如下：

library(jiebaR)
library(tidyverse)
library(rvest)
library(wordcloud2)

url <- "http://www.gov.cn/premier/2020-05/22/content_5513757.htm"
text <- url %>%
  read_html(encoding="utf-8") %>%
  html_nodes("div.pages_content") %>%
  html_text()

stopwords_CN <- c("被","怎么","还是","多少","得", "吗","给",
                  "年","月","还","个","能", "日","什么","做","没","啊",
                  "的", "了", "在", "是", "我", "有", "和", "就","不",
                  "人", "都", "一", "一个", "上", "也", "很", "到", "说",
                  "要", "去", "你","会", "着", "没有", "看", "好",
                  "自己", "这", "等")

#初始化分词引擎
engine <- worker()

#开始分词
seg <- segment(text,engine)
tibble(seg) %>%
  count(seg,sort=TRUE) %>%
  filter(!seg %in% stopwords_CN) %>%
  top_n(150) %>%
  wordcloud2(size=1,shape="star")

生成的词云如下：

参考资料：

[1]

dragon25428的github仓库资料: https://github.com/dragon25428/R-Web-Crawler/blob/master/政府报告词云图.R

[2]

政府工作报告（文字实录）: http://www.gov.cn/premier/2020-05/22/content_5513757.htm