R文本分析

jiebaR

---
title: "景区点评分析"
output: word_document
---


#读入点评内容数据
#install.packages("RMySQL")

#library(RODBC)
#myconn<-odbcConnect("wuchaojin",uid = "root",pwd = "admin")
#scenic<-sqlQuery(myconn,"select 景区名称 from dianping")
#comment_time<-sqlQuery(myconn,"select date_format(点评时间,'%Y-%m') from #dianping order by 1")
#pundat<-sqlQuery(myconn,"select 点评内容 from dianping")
#jiebaR
library(jiebaR)
mixseg = worker("tag",stop_word="C:\\Users\\wuchaojin\\Desktop\\stopword.txt")
#分词
split_word<-mixseg[as.character(pundat[,1])]
#词频
split_word<-freq(split_word)     #词频
split_word<-as.data.frame(split_word,responseName="freq")    #转为数据框
split_word<-split_word[order(-split_word$freq),]    #按词频大小排序
#插入分词类别




canyin<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\餐饮.csv")
fengjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\风景景点.csv")
fuwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\服务.csv")
ruyuan<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\购票入园.csv")
gouwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\购物.csv")
guanli<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\管理.csv")
huanjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\环境.csv")
jiage<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\价格.csv")
jiaotong<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\交通.csv")
sheshi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\设施设备.csv")
tianqi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\天气.csv")
zhusu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\名镇古迹\\住宿.csv")

split_word$category[split_word$char %in% canyin$word ]<-"餐饮"
split_word$category[split_word$char %in% fengjing$word ]<-"风景景点"
split_word$category[split_word$char %in% fuwu$word ]<-"服务"
split_word$category[split_word$char %in% ruyuan$word ]<-"购票入园"
split_word$category[split_word$char %in% gouwu$word ]<-"购物"
split_word$category[split_word$char %in% guanli$word ]<-"管理"
split_word$category[split_word$char %in% huanjing$word ]<-"环境"
split_word$category[split_word$char %in% jiage$word ]<-"价格"
split_word$category[split_word$char %in% jiaotong$word ]<-"交通"
split_word$category[split_word$char %in% sheshi$word ]<-"设施"
split_word$category[split_word$char %in% tianqi$word ]<-"天气"
split_word$category[split_word$char %in% zhusu$word ]<-"住宿"


#数据透视表-各类别词频
freq_data<-aggregate(split_word$freq,by=list(split_word$category),FUN=sum,na.rm=TRUE)
#变量重命名
library(reshape)
freq_data_all<-rename(freq_data,c(Group.1="category",x="freq_all"))
freq_data_all<-freq_data_all[order(-freq_data_all$freq_all),]
#输出外部文件
write.csv(split_word,"new_split_all.csv",row.names = FALSE)  #分词词频及类别列表
write.csv(freq_data_all,"freq_data_all.csv",row.names = FALSE)  #各维度词频数


#一星


library(jiebaR)
mixseg1 = worker("tag",stop_word="C:\\Users\\wuchaojin\\Desktop\\stopword.txt")
split_word1<-mixseg1[as.character(pundat1[,1])]
split_word1<-freq(split_word1)     #词频
split_word1<-as.data.frame(split_word1,responseName="freq")    #转为数据框
split_word1<-split_word1[order(-split_word1$freq),]    #按词频大小排序
canyin<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\canyin.csv")
fuwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\fuwu.csv")
huanjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\huanjing.csv")
jiage<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiage.csv")
jiaotong<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiaotong.csv")
guanli<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\guanli.csv")
sheshi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\sheshi.csv")
ruyuan<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\ruyuan.csv")
gouwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\gouwu.csv")
zhusu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\zhusu.csv")
tianqi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\tianqi.csv")
xiangmu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\xiangmu.csv")


split_word1$category[split_word1$char %in% tianqi$word ]<-"天气"
split_word1$category[split_word1$char %in% canyin$word ]<-"餐饮"
split_word1$category[split_word1$char %in% fuwu$word ]<-"服务"
split_word1$category[split_word1$char %in% xiangmu$word ]<-"娱乐项目"
split_word1$category[split_word1$char %in% jiage$word ]<-"价格"
split_word1$category[split_word1$char %in% jiaotong$word ]<-"交通"
split_word1$category[split_word1$char %in% sheshi$word ]<-"设施"
split_word1$category[split_word1$char %in% ruyuan$word ]<-"购票入园"
split_word1$category[split_word1$char %in% guanli$word ]<-"管理"
split_word1$category[split_word1$char %in% zhusu$word ]<-"住宿"
split_word1$category[split_word1$char %in% huanjing$word ]<-"环境氛围"
split_word1$category[split_word1$char %in% gouwu$word ]<-"购物"
freq_data1<-aggregate(split_word1$freq,by=list(split_word1$category),FUN=sum,na.rm=TRUE)
library(reshape)
freq_data_1<-rename(freq_data1,c(Group.1="category",x="freq_1"))

write.csv(split_word1,"new_split_1.csv",row.names = FALSE)
write.csv(freq_data_1,"freq_data_1.csv",row.names = FALSE)


#二星


library(jiebaR)
mixseg2 = worker("tag",stop_word="C:\\Users\\wuchaojin\\Desktop\\stopword.txt")
split_word2<-mixseg2[as.character(pundat2[,1])]
split_word2<-freq(split_word2)     #词频
split_word2<-as.data.frame(split_word2,responseName="freq")    #转为数据框
split_word2<-split_word2[order(-split_word2$freq),]    #按词频大小排序
canyin<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\canyin.csv")
fuwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\fuwu.csv")
huanjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\huanjing.csv")
jiage<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiage.csv")
jiaotong<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiaotong.csv")
guanli<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\guanli.csv")
sheshi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\sheshi.csv")
ruyuan<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\ruyuan.csv")
gouwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\gouwu.csv")
zhusu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\zhusu.csv")
tianqi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\tianqi.csv")
xiangmu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\xiangmu.csv")


split_word2$category[split_word2$char %in% tianqi$word ]<-"天气"
split_word2$category[split_word2$char %in% canyin$word ]<-"餐饮"
split_word2$category[split_word2$char %in% fuwu$word ]<-"服务"
split_word2$category[split_word2$char %in% xiangmu$word ]<-"娱乐项目"
split_word2$category[split_word2$char %in% jiage$word ]<-"价格"
split_word2$category[split_word2$char %in% jiaotong$word ]<-"交通"
split_word2$category[split_word2$char %in% sheshi$word ]<-"设施"
split_word2$category[split_word2$char %in% ruyuan$word ]<-"购票入园"
split_word2$category[split_word2$char %in% guanli$word ]<-"管理"
split_word2$category[split_word2$char %in% zhusu$word ]<-"住宿"
split_word2$category[split_word2$char %in% huanjing$word ]<-"环境氛围"
split_word2$category[split_word2$char %in% gouwu$word ]<-"购物"
freq_data2<-aggregate(split_word2$freq,by=list(split_word2$category),FUN=sum,na.rm=TRUE)
library(reshape)
freq_data_2<-rename(freq_data2,c(Group.1="category",x="freq_2"))
write.csv(split_word2,"new_split_2.csv",row.names = FALSE)
write.csv(freq_data_2,"freq_data_2.csv",row.names = FALSE)


#三星

library(jiebaR)
mixseg3 = worker("tag",stop_word="C:\\Users\\wuchaojin\\Desktop\\stopword.txt")
split_word3<-mixseg3[as.character(pundat3[,1])]
split_word3<-freq(split_word3)     #词频
split_word3<-as.data.frame(split_word3,responseName="freq")    #转为数据框
split_word3<-split_word3[order(-split_word3$freq),]    #按词频大小排序
canyin<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\canyin.csv")
fuwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\fuwu.csv")
huanjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\huanjing.csv")
jiage<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiage.csv")
jiaotong<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiaotong.csv")
guanli<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\guanli.csv")
sheshi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\sheshi.csv")
ruyuan<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\ruyuan.csv")
gouwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\gouwu.csv")
zhusu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\zhusu.csv")
tianqi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\tianqi.csv")
xiangmu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\xiangmu.csv")


split_word3$category[split_word3$char %in% tianqi$word ]<-"天气"
split_word3$category[split_word3$char %in% canyin$word ]<-"餐饮"
split_word3$category[split_word3$char %in% fuwu$word ]<-"服务"
split_word3$category[split_word3$char %in% xiangmu$word ]<-"娱乐项目"
split_word3$category[split_word3$char %in% jiage$word ]<-"价格"
split_word3$category[split_word3$char %in% jiaotong$word ]<-"交通"
split_word3$category[split_word3$char %in% sheshi$word ]<-"设施"
split_word3$category[split_word3$char %in% ruyuan$word ]<-"购票入园"
split_word3$category[split_word3$char %in% guanli$word ]<-"管理"
split_word3$category[split_word3$char %in% zhusu$word ]<-"住宿"
split_word3$category[split_word3$char %in% huanjing$word ]<-"环境氛围"
split_word3$category[split_word3$char %in% gouwu$word ]<-"购物"
freq_data3<-aggregate(split_word3$freq,by=list(split_word3$category),FUN=sum,na.rm=TRUE)
library(reshape)
freq_data_3<-rename(freq_data3,c(Group.1="category",x="freq_3"))
write.csv(split_word3,"new_split_3.csv",row.names = FALSE)
write.csv(freq_data_3,"freq_data_3.csv",row.names = FALSE)

#四星

library(jiebaR)
mixseg4 = worker("tag",stop_word="C:\\Users\\wuchaojin\\Desktop\\stopword.txt")
split_word4<-mixseg4[as.character(pundat4[,1])]
split_word4<-freq(split_word4)     #词频
split_word4<-as.data.frame(split_word4,responseName="freq")    #转为数据框
split_word4<-split_word4[order(-split_word4$freq),]    #按词频大小排序
canyin<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\canyin.csv")
fuwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\fuwu.csv")
huanjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\huanjing.csv")
jiage<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiage.csv")
jiaotong<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiaotong.csv")
guanli<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\guanli.csv")
sheshi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\sheshi.csv")
ruyuan<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\ruyuan.csv")
gouwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\gouwu.csv")
zhusu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\zhusu.csv")
tianqi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\tianqi.csv")
xiangmu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\xiangmu.csv")


split_word4$category[split_word4$char %in% tianqi$word ]<-"天气"
split_word4$category[split_word4$char %in% canyin$word ]<-"餐饮"
split_word4$category[split_word4$char %in% fuwu$word ]<-"服务"
split_word4$category[split_word4$char %in% xiangmu$word ]<-"娱乐项目"
split_word4$category[split_word4$char %in% jiage$word ]<-"价格"
split_word4$category[split_word4$char %in% jiaotong$word ]<-"交通"
split_word4$category[split_word4$char %in% sheshi$word ]<-"设施"
split_word4$category[split_word4$char %in% ruyuan$word ]<-"购票入园"
split_word4$category[split_word4$char %in% guanli$word ]<-"管理"
split_word4$category[split_word4$char %in% zhusu$word ]<-"住宿"
split_word4$category[split_word4$char %in% huanjing$word ]<-"环境氛围"
split_word4$category[split_word4$char %in% gouwu$word ]<-"购物"
freq_data4<-aggregate(split_word4$freq,by=list(split_word4$category),FUN=sum,na.rm=TRUE)
library(reshape)
freq_data_4<-rename(freq_data4,c(Group.1="category",x="freq_4"))
write.csv(split_word4,"new_split_4.csv",row.names = FALSE)
write.csv(freq_data_4,"freq_data_4.csv",row.names = FALSE)


#五星

library(jiebaR)
mixseg5 = worker("tag",stop_word="C:\\Users\\wuchaojin\\Desktop\\stopword.txt")
split_word5<-mixseg5[as.character(pundat5[,1])]
split_word5<-freq(split_word5)     #词频
split_word5<-as.data.frame(split_word5,responseName="freq")    #转为数据框
split_word5<-split_word5[order(-split_word5$freq),]    #按词频大小排序
canyin<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\canyin.csv")
fuwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\fuwu.csv")
huanjing<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\huanjing.csv")
jiage<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiage.csv")
jiaotong<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\jiaotong.csv")
guanli<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\guanli.csv")
sheshi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\sheshi.csv")
ruyuan<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\ruyuan.csv")
gouwu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\gouwu.csv")
zhusu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\zhusu.csv")
tianqi<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\tianqi.csv")
xiangmu<-read.csv("C:\\Users\\wuchaojin\\Desktop\\category\\xiangmu.csv")


split_word5$category[split_word5$char %in% tianqi$word ]<-"天气"
split_word5$category[split_word5$char %in% canyin$word ]<-"餐饮"
split_word5$category[split_word5$char %in% fuwu$word ]<-"服务"
split_word5$category[split_word5$char %in% xiangmu$word ]<-"娱乐项目"
split_word5$category[split_word5$char %in% jiage$word ]<-"价格"
split_word5$category[split_word5$char %in% jiaotong$word ]<-"交通"
split_word5$category[split_word5$char %in% sheshi$word ]<-"设施"
split_word5$category[split_word5$char %in% ruyuan$word ]<-"购票入园"
split_word5$category[split_word5$char %in% guanli$word ]<-"管理"
split_word5$category[split_word5$char %in% zhusu$word ]<-"住宿"
split_word5$category[split_word5$char %in% huanjing$word ]<-"环境氛围"
split_word5$category[split_word5$char %in% gouwu$word ]<-"购物"
freq_data5<-aggregate(split_word5$freq,by=list(split_word5$category),FUN=sum,na.rm=TRUE)
library(reshape)
freq_data_5<-rename(freq_data5,c(Group.1="category",x="freq_5"))
write.csv(split_word5,"new_split_5.csv",row.names = FALSE)
write.csv(freq_data_5,"freq_data_5.csv",row.names = FALSE)



freq_data_all$freq_1<-freq_data_1[match(freq_data_all$category,freq_data_1$category),2]
freq_data_all$freq_2<-freq_data_2[match(freq_data_all$category,freq_data_2$category),2]
freq_data_all$freq_3<-freq_data_3[match(freq_data_all$category,freq_data_3$category),2]
freq_data_all$freq_4<-freq_data_4[match(freq_data_all$category,freq_data_4$category),2]
freq_data_all$freq_5<-freq_data_5[match(freq_data_all$category,freq_data_5$category),2]

freq_data_all$freq_1[is.na(freq_data_all$freq_1)]<-0
freq_data_all$freq_2[is.na(freq_data_all$freq_2)]<-0
freq_data_all$freq_3[is.na(freq_data_all$freq_3)]<-0
freq_data_all$freq_4[is.na(freq_data_all$freq_4)]<-0
freq_data_all$freq_5[is.na(freq_data_all$freq_5)]<-0

total<-freq_data_all
total_amount<-total
total_level<-transform(total,level=1*freq_1/freq_all+2*freq_2/freq_all+3*freq_3/freq_all+4*freq_4/freq_all+5*freq_5/freq_all)
total_l_a<-merge(total_amount[,c(1,2)],total_level[c(1,8)],by="category")
total_l_a<-total_l_a[order(-total_l_a$freq_all),]
#write.csv(total_level,"total_level.csv",row.names = FALSE)
#write.csv(total_amount,"total_amount.csv",row.names = FALSE)
#write.csv(total_l_a,"total_l_a.csv",row.names = FALSE)





## 一、研究目的

&#8195;&#8195;为充分了解现代旅游者决策、购买、游览和游后行为，洞察游客反馈及潜在游客需求，对游客在网上分享游览经验，交流出游方式、行程路线、交通工具、风土人情、景区景点、权益维护、态度评价和心得体会等方面信息的研究变得至关重要。网络*开放共享的特性可充分反映游客对景区形象的认知和感受，这里以主要OTA点评数据为样本，使用R语言提取游客对`r scenic[1,1]`旅游形象感知和评价的高频特征词，研究游客对景区各维度的关注度及满意度，挖掘游客潜在需求，为景区形象的完善与提升提供依据。

## 二、研究方法

#### 2.1数据获取
&#8195;&#8195;运用爬虫工具采集景区点评数据，并存入本地数据库，采集数据包括：景区名称、总点评数、每条点评内容、点评时间、点评星级，最后设置格式并清洗数据。

#### 2.2分词处理
&#8195;&#8195;连接数据库提取点评数据，载入分词包jiebaR进行分词，按词性筛选并剔除无用停用词，然后做词频分析。

#### 2.3词库整理
&#8195;&#8195;对分词进行分类，参考该景区主题类型，根据点评分词涉及范围，将景区分成各个维度，得出景区维度：如：娱乐项目、餐饮、服务、环境、价格、交通、设施、入园、管理等多个分析维度。将各维度涉及分词整理成词库，以备后用。并将该景区相关专用名词加入jiebaR下面的分词词典，以重新获得更准确的分词。整理好的词库可适用于同主题同类景区,便于批量生成报告。

#### 2.4满意度分析
&#8195;&#8195;按星级分类统计词频，调取各维度词库分词，加权平均求各维度满意度指数。

#### 2.5图形输出
&#8195;&#8195;按点评时间输出趋势图衡量景区热度及满意度走势，按景区维度输出分布图研究游客潜在需求及对景区感知，最后输出词云反映实际具体问题。







overall_level1$星级<-as.numeric(overall_level1$星级)
overall_level2<-round(mean(overall_level1$星级)/5,2)
overall_level3<-aggregate(overall_level1$星级,by=list(overall_level1$时间),FUN=mean,na.rm=TRUE)
library(reshape)
overall_level3<-rename(overall_level3,c(Group.1="时间",x="星级"))
#write.csv(overall_amount,"overall_amount.csv",row.names = FALSE)
#write.csv(overall_level3,"overall_level.csv",row.names = FALSE)



library(ggplot2)  
library(gtable)  
library(grid) 
#柱形图
#overall_amount<-read.csv("D:\\Documents\\overall_amount.csv",row.names = NULL)



## 三、研究结果
#### 3.1总体趋势
&#8195;&#8195;`r scenic[1,1]`从`r comment_time[1,1]`至今，主要OTA总点评数`r table(scenic)`，通过如下趋势图可看出，（在每年的XX-XX月份景区热度达到峰值，15年高峰期与14年相比同期增长XX%，16年相比15年同期增长XX%，景区整体热度每年呈XX趋势。）

景区关注度趋势图数据：

overall_amount



ggplot(overall_amount,aes(x=factor(时间),y=点评数))+
  geom_bar(stat="identity",fill="#BED742")+
  geom_text(aes(label=点评数),vjust=-0.4,colour="black",size=3)+
  ylim(0,max(overall_amount$点评数)*1.05)+
  theme_bw()+
  labs(x="日期",y="点评数")+  
  ggtitle(paste("景区关注度趋势图    总点评数：",comment_num,sep=""))+  
  theme(axis.text.x=element_text(angle=65,hjust=1,vjust=1,size=8),
        axis.title.x=element_text(colour="darkgrey",size=10),
        axis.title.y=element_text(colour="darkgrey",size=10),
        panel.border=element_blank(),  
        panel.grid.major.x=element_blank(),  
        panel.grid.minor.x=element_blank(),  
        panel.grid.major.y=element_blank(),  
        panel.grid.minor.y=element_blank(),
        axis.line=element_line(color="gray",size=1))



&#8195;&#8195;景区总体满意度达`r round(mean(overall_level1$星级)/5,2)`（0~1），（在出游高峰平均满意度平均水平为XX。15年出游高峰满意度同比增长XX%，16年满意度相比15年同期增长XX%，景区整体满意度逐年XX。）

景区满意度趋势图数据：

#overall_level<-read.csv("D:\\Documents\\overall_level.csv")
overall_level3

#折线图
library(ggplot2)
ggplot(overall_level3,aes(x=factor(时间),y=星级/5,group=1))+
  geom_line(color="#BED742")+geom_point(color="#BED742")+
  geom_text(aes(label=round(星级/5,2)),vjust=-0.4,colour="black",size=3)+
  ylim(min(overall_level3$星级/5)*0.9,max(overall_level3$星级/5)*1.05)+
 theme_bw()+
  labs(x="日期",y="满意度")+  
  ggtitle(paste("景区满意度趋势图    总体满意度",round(mean(overall_level1$星级)/5,2),sep=""))+
  theme(axis.text.x=element_text(angle=65,hjust=1,vjust=1,size=8),
        axis.title.x=element_text(colour="darkgrey",size=9),
        axis.title.y=element_text(colour="darkgrey",size=9),
        panel.border=element_blank(),  
        panel.grid.major.x=element_blank(),  
        panel.grid.minor.x=element_blank(),  
        panel.grid.major.y=element_blank(),  
        panel.grid.minor.y=element_blank())


#### 3.2 维度细分

  通过下图景区各维度满意度分布，（可见游客对景区的XX最为关注，平均每条评论提及XX次。其次是XX、XX以及XX，人均提及近XX次。这充分反映了游客对该景区以及该类景区的热切关注点以及潜在需求点。合理丰富娱乐项目，加强园内设施及管理，合理设置调整门票及园内项目价格，重视园内餐饮，根据顾客需求有针对性发展及创新。）



景区各维度关注度及满意度数据：

#total_data<-read.csv("D:\\Documents\\total_l_a.csv")
total_l_a



library(gcookbook)
library(ggplot2)


ggplot(total_l_a,aes(x=reorder(category,-freq_all),y=freq_all/table(scenic)))+
  geom_bar(stat="identity",fill="#BED742")+
  geom_text(aes(label=round(total_l_a$freq_all/table(scenic),2)),vjust=-0.4,colour="black",size=4)+
  ylim(0,max(total_l_a$freq_all/table(scenic))*1.1)+
  theme_bw()+
  labs(x="景区维度",y="平均单条点评提及频次")+  
  ggtitle("景区各维度关注度")+
  theme(axis.text.x=element_text(angle=60,hjust=1,vjust=1,size=10),
        axis.title.x=element_text(colour="darkgrey",size=10),
        axis.title.y=element_text(colour="darkgrey",size=10),
        panel.border=element_blank(),  
        panel.grid.major.x=element_blank(),  
        panel.grid.minor.x=element_blank(),  
        panel.grid.major.y=element_blank(),  
        panel.grid.minor.y=element_blank(),
        axis.line.y=element_line(color="gray",size=1),
        axis.line.x=element_line(color="gray",size=1))






&#8195;&#8195;(景区各维度满意度整体分布在XX-XX之间，其中，对娱乐项目关注度最高且满意度也较高达XX，管理及设施方面游客也极为关注但满意度较低，景区应提起对这两方面的重视，选择合理化的园内管理方式，改善园内设施，环境卫生以及服务质量用户体验较差，急需提升。)



ggplot(total_l_a,aes(reorder(x=category,-freq_all),y=level/5,group=1))+geom_line(color="#BED742")+
  geom_text(aes(label=round(level/5,2)),vjust=-0.4,colour="black",size=3.5)+
  ylim(min(total_l_a$level/5)*0.95,max(total_l_a$level/5)*1.05)+
  geom_point(color="#BED742")+
  theme_bw()+
  labs(x="景区维度",y="满意度")+  
  ggtitle("景区各维度满意度")+
  theme(axis.text.x=element_text(angle=60,hjust=1,vjust=1,size=10),
        axis.title.x=element_text(colour="darkgrey",size=10),
        axis.title.y=element_text(colour="darkgrey",size=10),
        panel.border=element_blank(),  
        panel.grid.major.x=element_blank(),  
        panel.grid.minor.x=element_blank(),  
        panel.grid.major.y=element_blank(),  
        panel.grid.minor.y=element_blank())


#### 3.3 具体问题反馈

&#8195;&#8195;通过词云展现了点评高频词汇，可直观展现游客所关注的问题有哪些，(娱乐项目维度有热带风暴、大喇叭、深海巨蟒、滑道等比较受关注，管理方面排队问题备受关注，拖鞋、储物柜、更衣室、柜子等词出现频繁，说明储物设施方面也应着重管理等。)


data<-read.csv("D:\\Documents\\new_split_all.csv")
library(wordcloud)


head(data)


wordcloud(data$char,data$freq,min.freq=max(data$freq)*0.03,random.order=FALSE,random.color=TRUE,
          colors=c('pink','red','blue','green','yellow','purple','beige','brown','peru','khaki'))



&#8195;&#8195;下面是针对一星差评点评做出的词云统计，(去除整体点评用户均会涉及的“项目”，可见“排队”，“插队”， “人多”， “服务“，“工作人员”，“垃圾”，“脏”，“恶心”，“设施”等词汇在差评中出现频次颇多，)在一定程度上反映了一些急需解决的实际问题，同时也印证了3.2中各景区维度满意度的评估。


data1<-read.csv("D:\\Documents\\new_split_1.csv")



wordcloud(data1$char,data$freq,min.freq=max(data$freq)*0.001,max.words=50,random.order=FALSE,random.color=TRUE,
          colors=c('pink','red','blue','green','yellow','purple','beige','brown','peru','khaki'))

Rwordseg

reviewpath <- "D:/work/桌面/点评报告/情感分析" 
completepath <- list.files(reviewpath, pattern = "*.txt$", full.names = TRUE)

####批量读入文本  
read.txt <- function(x) {  
  des <- readLines(x)                   #每行读取  
  return(paste(des, collapse = ""))     #没有return则返回最后一个函数对象  
}  
review <- lapply(completepath, read.txt)  
#如果程序警告，这里可能是部分文件最后一行没有换行导致，不用担心
#中文主要有知网整理的情感词典Hownet和*大学整理发布的NTUSD两个情感词典，还有哈工大信息检索研究室开源的《同义词词林》可以用于情感词典的扩充。

setwd("D:\\work\\桌面\\点评报告\\情感分析\\中文情感极性词典")
library(openxlsx)
posneg<-read.xlsx("posneg.xlsx",1)

dict<-posneg[,"term"]  

install.packages("Rwordseg", repos = "http://R-Forge.R-project.org")
library(Rwordseg)  
#listDict()  #查看已有词库  
#uninstallDict() #删除安装的词典    
insertWords(dict)  


train.test<-read.xlsx("D:\\work\\桌面\\点评报告\\情感分析\\test1.xlsx",1)

sentence <- as.vector(train.test$点评内容) #文本内容转化为向量sentence  
sentence <- gsub("[[:digit:]]*", "", sentence) #清除数字[a-zA-Z]  
sentence <- gsub("[a-zA-Z]", "", sentence)   #清除英文字符  
sentence <- gsub("\\.", "", sentence)      #清除全英文的dot符号  
train.test <- train.test[!is.na(sentence), ]          #清除一些空值文本（文本名）  
sentence<- sentence[!is.na(sentence)]   #清除对应sentence里面的空值（文本内容），要先执行文本名  
train.test<- train.test[!nchar(sentence) < 2, ]  #筛选字符数小于2的文本  
sentence<- sentence[!nchar(sentence) < 2] #`nchar`函数对字符计数，英文叹号为R语言里的“非”函数  

system.time(x <- segmentCN(strwords = sentence))   
#每次可能耗费时间较长的过程，都要使用少量数据预估一下时间，这是一个优秀的习惯  
temp <- lapply(x, length) #每一个元素的长度,即文本分出多少个词  

head(testterm,100)
head(x)

temp<- unlist(temp)  #lapply返回的是一个list，所以3行unlist  
id <- rep(test[, "id"], temp) #将每一个对应的id复制相应的次数，就可以和词汇对应了  
label <- rep(test[, "label"], temp)#id对应的情感倾向标签复制相同的次数  
term <- unlist(x) #将list解散为向量  
testterm <- as.data.frame(cbind(id, term, label), stringsAsFactors = F)  
#将一一对应的三个向量按列捆绑为数据框，分词整理就基本结束了  

x<- unlist(x)
x<-data.frame(x)
names(x)<-'term'
library(plyr)  
testterm <- join(x, posneg)  
testterm <- testterm[!is.na(testterm$weight), ]  
head(testterm)



segmentCN(strwords = "地址是圣诺亚大厦",returnType="tm")