商品评论情感分析之思路分析

import jieba
words = jieba.cut("从编程小白的角度看，入门极佳")

from sklearn.naive_bayes import MultinomialNB
nb = MultinomialNB(alpha=1.0)  # alpha是拉普拉斯平滑系数

import pandas as pd
import numpy as np
import jieba
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.naive_bayes import MultinomialNB
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

# 1. 模拟商品评论数据
print("=== 模拟商品评论数据 ===")
data = pd.DataFrame({
    'content': [
        '从编程小白的角度看，入门极佳，推荐购买',
        '质量很好，用起来很舒服，好评',
        '性价比高，值得入手，非常满意',
        '物流快，包装好，质量不错',
        '差评，质量太差了，浪费钱',
        '不好用，太失望了，不推荐',
        '太差了，退货了，质量堪忧',
        '失望，与描述不符，差评'
    ],
    'sentiment': ['好评', '好评', '好评', '好评', '差评', '差评', '差评', '差评']
})

# 新增labels列
data['labels'] = data['sentiment'].map({'好评': 1, '差评': 0})
print(data)

# 2. 分词
print("\n=== 分词 ===")

def cut_words(text):
    """使用jieba分词"""
    words = jieba.cut(text)
    return ' '.join(words)

# 对所有评论进行分词
data['cut_content'] = data['content'].apply(cut_words)

for _, row in data.iterrows():
    print(f"原文: {row['content']}")
    print(f"分词: {row['cut_content']}")
    print()

# 3. 删除无效词
print("=== 删除无效词 ===")

# 定义无效词列表
stop_words = ['，', '。', '、', '了', '的', '从', '看', '很', '太']

def remove_stop_words(text):
    """删除无效词"""
    words = text.split()
    valid_words = [word for word in words if word not in stop_words]
    return ' '.join(valid_words)

data['clean_content'] = data['cut_content'].apply(remove_stop_words)

for _, row in data.iterrows():
    print(f"分词后: {row['cut_content']}")
    print(f"清理后: {row['clean_content']}")
    print()

# 4. 特征提取
print("=== 特征提取 ===")

vectorizer = CountVectorizer()
X = vectorizer.fit_transform(data['clean_content'])
print(f"特征矩阵形状: {X.shape}")
print(f"特征名称: {vectorizer.get_feature_names_out()}")

# 5. 划分数据集
print("\n=== 划分数据集 ===")
y = data['labels']
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.25, random_state=42
)
print(f"训练集: {X_train.shape}")
print(f"测试集: {X_test.shape}")

# 6. 训练模型
print("\n=== 训练模型 ===")
nb = MultinomialNB(alpha=1.0)
nb.fit(X_train, y_train)

# 7. 评估模型
print("\n=== 评估模型 ===")
y_pred = nb.predict(X_test)
acc = accuracy_score(y_test, y_pred)
print(f"准确率: {acc:.4f}")

# 8. 预测新评论
print("\n=== 预测新评论 ===")
new_comments = [
    '质量很好，推荐购买',
    '太差了，浪费钱'
]

for comment in new_comments:
    # 分词
    cut_comment = cut_words(comment)
    # 删除无效词
    clean_comment = remove_stop_words(cut_comment)
    # 向量化
    X_new = vectorizer.transform([clean_comment])
    # 预测
    pred = nb.predict(X_new)[0]
    proba = nb.predict_proba(X_new)[0]
    
    label = '好评' if pred == 1 else '差评'
    print(f"评论: {comment}")
    print(f"分词: {cut_comment}")
    print(f"清理: {clean_comment}")
    print(f"预测: {label} (好评概率: {proba[1]:.4f})")
    print()

# 9. 完整流程函数
def sentiment_analysis_pipeline():
    """情感分析完整流程"""
    print("=" * 50)
    print("商品评论情感分析")
    print("=" * 50)
    
    # 1. 读数据
    print("\n1. 读数据")
    data = pd.DataFrame({
        'content': [
            '从编程小白的角度看，入门极佳，推荐购买',
            '质量很好，用起来很舒服，好评',
            '性价比高，值得入手，非常满意',
            '物流快，包装好，质量不错',
            '差评，质量太差了，浪费钱',
            '不好用，太失望了，不推荐',
            '太差了，退货了，质量堪忧',
            '失望，与描述不符，差评'
        ],
        'labels': [1, 1, 1, 1, 0, 0, 0, 0]
    })
    print(f"   数据量: {len(data)}")
    
    # 2. 分词
    print("\n2. 分词")
    data['cut_content'] = data['content'].apply(cut_words)
    
    # 3. 删除无效词
    print("\n3. 删除无效词")
    data['clean_content'] = data['cut_content'].apply(remove_stop_words)
    
    # 4. 特征提取
    print("\n4. 特征提取")
    vectorizer = CountVectorizer()
    X = vectorizer.fit_transform(data['clean_content'])
    y = data['labels']
    print(f"   特征数: {X.shape[1]}")
    
    # 5. 划分数据集
    print("\n5. 划分数据集")
    X_train, X_test, y_train, y_test = train_test_split(
        X, y, test_size=0.25, random_state=42
    )
    
    # 6. 训练模型
    print("\n6. 训练模型")
    nb = MultinomialNB(alpha=1.0)
    nb.fit(X_train, y_train)
    
    # 7. 评估
    print("\n7. 评估模型")
    y_pred = nb.predict(X_test)
    acc = accuracy_score(y_test, y_pred)
    print(f"   准确率: {acc:.4f}")
    
    return nb, vectorizer

# 运行完整流程
nb, vectorizer = sentiment_analysis_pipeline()