import pandas as pd
import jieba
from collections import Counter
import matplotlib.pyplot as plt
from wordcloud import WordCloud
from snownlp import SnowNLP
import matplotlib
matplotlib.use('Agg') # 兼容无弹窗环境，CodeGeeX网页版必备
解决中文乱码
plt.rcParams["font.family"] = ["SimHei", "WenQuanYi Micro Hei", "Heiti TC"]
plt.rcParams["axes.unicode_minus"] = False
自动兼容 utf-8/gbk 两种CSV编码
try:
df = pd.read_csv("yibin_comment.csv", encoding="utf-8")
except Exception:
df = pd.read_csv("yibin_comment.csv", encoding="gbk")
清洗空评论、多余空格
df["comment"] = df["comment"].astype(str).str.strip()
df = df[df["comment"] != ""]
comments = df["comment"]
SnowNLP情感打分分类
sentiment_result = []
for text in comments:
s = SnowNLP(text)
score = s.sentiments
if score > 0.5:
sentiment_result.append("正面")
elif score < 0.5:
sentiment_result.append("负面")
else:
sentiment_result.append("中性")
df["情感类别"] = sentiment_result
统计情感数量与占比
stat = df["情感类别"].value_counts()
print("===== 宜宾旅游口碑情感统计 =====")
print(stat)
print("\n各类占比：")
print((stat / len(df)).round(4))
绘制情感饼图并保存
plt.figure(figsize=(8, 6))
plt.pie(stat.values, labels=stat.index, autopct="%1.2f%%", shadow=True, startangle=90)
plt.title("宜宾（蜀南竹海/李庄/五粮液）游客情感分布饼图")
plt.axis("equal")
plt.savefig("情感饼图.png", dpi=300, bbox_inches="tight")
plt.close()
分词、过滤停用词与单字
stop_words = {"的", "了", "很", "也", "有", "比较", "一点", "还", "就", "都", "感觉", "里面", "可以",
"过来", "就是", "在", "和", "我", "去", "到", "一个", "这个", "那个", "不", "没有", "太"}
all_words = []
for text in comments:
words = jieba.cut(text)
for w in words:
if len(w) > 1 and w not in stop_words:
all_words.append(w)
word_count = Counter(all_words)
top_words = word_count.most_common(35)
word_str = " ".join([w[0] for w in top_words])
print("\n===== Top35高频关键词 =====")
for word, num in top_words:
print(f"{word}: {num}")
生成词云（CodeGeeX网页环境大概率缺少simhei.ttf字体）
wc = WordCloud(
background_color="white",
font_path="simhei.ttf",
width=1000, height=600,
max_words=200
)
wc.generate(word_str)
plt.figure(figsize=(10, 6))
plt.imshow(wc, interpolation="bilinear")
plt.axis("off")
plt.title("宜宾旅游高频关键词词云图")
plt.savefig("关键词云图.png", dpi=300, bbox_inches="tight")
plt.close()
print("\n运行结束，已输出两张图片：情感饼图.png、关键词云图.png")
import pandas as pd
import jieba
from collections import Counter
import matplotlib.pyplot as plt
from wordcloud import WordCloud
from snownlp import SnowNLP
import matplotlib
matplotlib.use('Agg') # 兼容无弹窗环境，CodeGeeX网页版必备
解决中文乱码
plt.rcParams["font.family"] = ["SimHei", "WenQuanYi Micro Hei", "Heiti TC"]
plt.rcParams["axes.unicode_minus"] = False
自动兼容 utf-8/gbk 两种CSV编码
try:
df = pd.read_csv("yibin_comment.csv", encoding="utf-8")
except Exception:
df = pd.read_csv("yibin_comment.csv", encoding="gbk")
清洗空评论、多余空格
df["comment"] = df["comment"].astype(str).str.strip()
df = df[df["comment"] != ""]
comments = df["comment"]
SnowNLP情感打分分类
sentiment_result = []
for text in comments:
s = SnowNLP(text)
score = s.sentiments
if score > 0.5:
sentiment_result.append("正面")
elif score < 0.5:
sentiment_result.append("负面")
else:
sentiment_result.append("中性")
df["情感类别"] = sentiment_result
统计情感数量与占比
stat = df["情感类别"].value_counts()
print("===== 宜宾旅游口碑情感统计 =====")
print(stat)
print("\n各类占比：")
print((stat / len(df)).round(4))
绘制情感饼图并保存
plt.figure(figsize=(8, 6))
plt.pie(stat.values, labels=stat.index, autopct="%1.2f%%", shadow=True, startangle=90)
plt.title("宜宾（蜀南竹海/李庄/五粮液）游客情感分布饼图")
plt.axis("equal")
plt.savefig("情感饼图.png", dpi=300, bbox_inches="tight")
plt.close()
分词、过滤停用词与单字
stop_words = {"的", "了", "很", "也", "有", "比较", "一点", "还", "就", "都", "感觉", "里面", "可以",
"过来", "就是", "在", "和", "我", "去", "到", "一个", "这个", "那个", "不", "没有", "太"}
all_words = []
for text in comments:
words = jieba.cut(text)
for w in words:
if len(w) > 1 and w not in stop_words:
all_words.append(w)
word_count = Counter(all_words)
top_words = word_count.most_common(35)
word_str = " ".join([w[0] for w in top_words])
print("\n===== Top35高频关键词 =====")
for word, num in top_words:
print(f"{word}: {num}")
生成词云（CodeGeeX网页环境大概率缺少simhei.ttf字体）
wc = WordCloud(
background_color="white",
font_path="simhei.ttf",
width=1000, height=600,
max_words=200
)
wc.generate(word_str)
plt.figure(figsize=(10, 6))
plt.imshow(wc, interpolation="bilinear")
plt.axis("off")
plt.title("宜宾旅游高频关键词词云图")
plt.savefig("关键词云图.png", dpi=300, bbox_inches="tight")
plt.close()
print("\n运行结束，已输出两张图片：情感饼图.png、关键词云图.png")