Python 爬取社交网络评论数据并完成情感分析

import requests
from bs4 import BeautifulSoup
import pandas as pd
import time

# 配置请求头，模拟浏览器访问（避免被反爬识别）
headers = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
    "Accept-Language": "zh-CN,zh;q=0.9",
    "Referer": "https://www.example.com/"  # 替换为目标网站
}

# 定义爬取函数
def crawl_comments(base_url, page_num):
    """
    爬取指定页数的评论数据
    :param base_url: 评论页面基础URL
    :param page_num: 要爬取的页数
    :return: 评论列表
    """
    comments_list = []
    for page in range(1, page_num + 1):
        try:
            # 构造分页URL（不同网站分页参数不同，需根据实际调整）
            url = f"{base_url}?page={page}"
            response = requests.get(url, headers=headers, timeout=10)
            response.encoding = "utf-8"  # 统一编码，避免乱码
            
            # 解析页面
            soup = BeautifulSoup(response.text, "html.parser")
            # 定位评论节点（需根据目标网站HTML结构调整class/id）
            comment_nodes = soup.find_all("div", class_="comment-content")
            
            # 提取评论内容
            for node in comment_nodes:
                comment = node.get_text(strip=True)
                if comment:  # 过滤空评论
                    comments_list.append({"page": page, "comment": comment})
            
            print(f"第{page}页爬取完成，共获取{len(comment_nodes)}条评论")
            time.sleep(1)  # 加延迟，避免请求过快被封IP
            
        except Exception as e:
            print(f"第{page}页爬取失败：{str(e)}")
            continue
    
    return comments_list

# 执行爬取（替换为实际目标URL，此处为示例）
if __name__ == "__main__":
    target_url = "https://www.example.com/product/comments"  # 替换为真实评论页URL
    total_pages = 5  # 爬取5页评论
    comments_data = crawl_comments(target_url, total_pages)
    
    # 转为DataFrame，便于后续处理
    df = pd.DataFrame(comments_data)
    # 保存为CSV文件，避免重复爬取
    df.to_csv("social_comments.csv", index=False, encoding="utf-8-sig")
    print(f"爬取完成！共获取{len(df)}条评论，已保存至social_comments.csv")

import pandas as pd
from snownlp import SnowNLP
import matplotlib.pyplot as plt

# 设置中文字体，避免图表乱码
plt.rcParams["font.sans-serif"] = ["SimHei"]
plt.rcParams["axes.unicode_minus"] = False

# 读取爬取的评论数据
df = pd.read_csv("social_comments.csv", encoding="utf-8-sig")

# 定义情感分析函数
def analyze_sentiment(comment):
    """
    计算单条评论的情感得分
    :param comment: 评论文本
    :return: 情感得分、情感标签
    """
    try:
        s = SnowNLP(comment)
        score = s.sentiments  # 情感得分（0-1）
        # 根据得分划分标签
        if score >= 0.7:
            label = "正面"
        elif score <= 0.3:
            label = "负面"
        else:
            label = "中性"
        return score, label
    except Exception as e:
        print(f"情感分析失败：{comment} | 错误：{str(e)}")
        return 0.5, "中性"  # 异常评论标记为中性

# 批量处理评论
df["sentiment_score"] = df["comment"].apply(lambda x: analyze_sentiment(x)[0])
df["sentiment_label"] = df["comment"].apply(lambda x: analyze_sentiment(x)[1])

# 保存分析结果
df.to_csv("comments_sentiment.csv", index=False, encoding="utf-8-sig")
print("情感分析完成！结果已保存至comments_sentiment.csv")

# 统计情感分布
sentiment_count = df["sentiment_label"].value_counts()
print("\n情感分布统计：")
print(sentiment_count)

# 可视化情感分布
plt.figure(figsize=(8, 6))
sentiment_count.plot(kind="pie", autopct="%1.1f%%", colors=["#66b3ff", "#ff9999", "#99ff99"])
plt.title("社交网络评论情感分布")
plt.ylabel("")  # 隐藏y轴标签
plt.savefig("sentiment_distribution.png", dpi=300, bbox_inches="tight")
plt.show()