Menü
情感分析技能,分析用户评论情感倾向,评估痛点严重程度,提取关键词和主题
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Use this skill when debugging complex, multi-factor, or long-running bugs. It maintains a project-root /coe Markdown case file per bug case using a strict Chain-of-Evidence model with only Problem, Hypothesis, and Evidence nodes.
Best practices for Remotion - Video creation in React
Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for "shadcn init", "create an app with --preset", or "switch to --preset".
Use when the user needs to design, implement, audit, or govern Storybook stories for frontend components, design systems, page states, interaction tests, accessibility checks, visual baselines, or component-driven UI workflows.
Allows you to view the user's screen as well as several hours of history. Use when the user makes a reference to their recent work, for which it'd be helpful to see the screen. This skill MUST be used whenever you need to resolve ambiguity in a user request, where the user hasn't specified enough context to do the task. Examples include disambiguating the specific user/app/document/error the user is referring to. You must also use this skill if the user asks about any question regarding Chronicle or asks what you can see from the screen.
Use when a frontend redesign or legacy CSS refactor is blocked by global style pollution, selector collisions, unstable DOM reuse, or the need to migrate a page or module to a new UI without breaking the old system.
| name | sentiment-analysis |
| description | 情感分析技能,分析用户评论情感倾向,评估痛点严重程度,提取关键词和主题 |
通过分析用户评论和讨论的情感倾向,量化用户满意度和痛点严重程度。
正面关键词:
love, great, awesome, perfect, excellent, amazing,
fantastic, wonderful, best, recommend, helpful,
easy, simple, fast, smooth, reliable
负面关键词:
hate, terrible, awful, worst, horrible, useless,
disappointing, frustrating, annoying, slow, buggy,
crash, broken, waste, regret, avoid
强度词:
very, extremely, absolutely, completely, totally,
really, so, too, incredibly, unbelievably
SQL 实现:
-- 简单情感分类
SELECT
id,
quote,
CASE
WHEN LOWER(quote) LIKE '%love%' OR LOWER(quote) LIKE '%great%'
OR LOWER(quote) LIKE '%awesome%' OR LOWER(quote) LIKE '%perfect%'
THEN 'positive'
WHEN LOWER(quote) LIKE '%hate%' OR LOWER(quote) LIKE '%terrible%'
OR LOWER(quote) LIKE '%awful%' OR LOWER(quote) LIKE '%worst%'
THEN 'negative'
ELSE 'neutral'
END as sentiment,
CASE
WHEN LOWER(quote) LIKE '%very%' OR LOWER(quote) LIKE '%extremely%'
THEN 'high'
ELSE 'medium'
END as intensity
FROM user_quotes;
情感评分 (-5 到 +5):
计算方法:
1. 统计正面关键词数量 (P)
2. 统计负面关键词数量 (N)
3. 统计强度词数量 (I)
4. 评分 = (P - N) × (1 + I × 0.5)
示例:
"I absolutely love this app!"
- P = 1 (love)
- N = 0
- I = 1 (absolutely)
- 评分 = (1 - 0) × (1 + 1 × 0.5) = 1.5
"This is the worst app ever!"
- P = 0
- N = 1 (worst)
- I = 0
- 评分 = (0 - 1) × (1 + 0) = -1
| 痛点 | 情感强度 | 频率 | 持续性 | 影响范围 | 总分 | 等级 |
|------|---------|------|--------|---------|------|------|
| 电池焦虑 | -4.2 | 45 | 6个月 | 广泛 | 18 | 严重 |
| 存储满 | -3.8 | 32 | 12个月 | 广泛 | 16 | 严重 |
| 通知多 | -2.5 | 18 | 3个月 | 中等 | 8 | 中等 |
评分标准:
-- 痛点严重程度分析
SELECT
category,
COUNT(*) as frequency,
AVG(sentiment_score) as avg_sentiment,
DATEDIFF(MAX(created_at), MIN(created_at)) as duration_days,
COUNT(DISTINCT source) as source_diversity,
-- 综合评分
(COUNT(*) * 0.3 +
ABS(AVG(sentiment_score)) * 2 +
DATEDIFF(MAX(created_at), MIN(created_at)) / 30 * 0.5 +
COUNT(DISTINCT source) * 0.5) as severity_score
FROM user_quotes
WHERE sentiment_score < 0 -- 只看负面
GROUP BY category
ORDER BY severity_score DESC;
方法:
1. 分词 (去除停用词)
2. 统计词频
3. 按频率排序
4. 提取 Top 20-30
停用词列表:
the, a, an, and, or, but, in, on, at, to, for,
of, with, by, from, is, are, was, were, be, been
SQL 实现 (简化版):
-- 提取包含特定词的评论数
SELECT
'battery' as keyword,
COUNT(*) as mention_count,
COUNT(*) * 100.0 / (SELECT COUNT(*) FROM user_quotes) as percentage
FROM user_quotes
WHERE LOWER(quote) LIKE '%battery%'
UNION ALL
SELECT
'storage' as keyword,
COUNT(*) as mention_count,
COUNT(*) * 100.0 / (SELECT COUNT(*) FROM user_quotes) as percentage
FROM user_quotes
WHERE LOWER(quote) LIKE '%storage%';
常见主题:
主题分类:
SELECT
CASE
WHEN LOWER(quote) LIKE '%slow%' OR LOWER(quote) LIKE '%lag%'
OR LOWER(quote) LIKE '%crash%'
THEN 'performance'
WHEN LOWER(quote) LIKE '%missing%' OR LOWER(quote) LIKE '%need%'
OR LOWER(quote) LIKE '%want%'
THEN 'feature_request'
WHEN LOWER(quote) LIKE '%confusing%' OR LOWER(quote) LIKE '%complicated%'
THEN 'ux_issue'
WHEN LOWER(quote) LIKE '%expensive%' OR LOWER(quote) LIKE '%price%'
THEN 'pricing'
WHEN LOWER(quote) LIKE '%bug%' OR LOWER(quote) LIKE '%broken%'
THEN 'bug'
ELSE 'other'
END as theme,
COUNT(*) as count
FROM user_quotes
GROUP BY theme
ORDER BY count DESC;
| 竞品 | 平均情感 | 正面% | 中性% | 负面% | 主要抱怨 |
|------|---------|-------|-------|-------|---------|
| A | +1.2 | 45% | 35% | 20% | 价格贵 |
| B | -0.5 | 25% | 40% | 35% | 功能少 |
| C | +2.1 | 60% | 30% | 10% | 无 |
SQL 查询:
SELECT
competitor_name,
AVG(sentiment_score) as avg_sentiment,
SUM(CASE WHEN sentiment = 'positive' THEN 1 ELSE 0 END) * 100.0 / COUNT(*) as positive_pct,
SUM(CASE WHEN sentiment = 'neutral' THEN 1 ELSE 0 END) * 100.0 / COUNT(*) as neutral_pct,
SUM(CASE WHEN sentiment = 'negative' THEN 1 ELSE 0 END) * 100.0 / COUNT(*) as negative_pct
FROM app_reviews
GROUP BY competitor_name
ORDER BY avg_sentiment DESC;
-- 月度情感趋势
SELECT
strftime('%Y-%m', created_at) as month,
AVG(sentiment_score) as avg_sentiment,
COUNT(*) as review_count
FROM user_quotes
GROUP BY month
ORDER BY month;
可视化:
月份: 01 02 03 04 05 06
情感: +1.2 +0.8 -0.5 -1.2 -1.5 -1.8 (趋势恶化)
## iOS 18 发布对情感的影响
**发布前 (2024-08)**:
- 平均情感: +0.5
- 负面评论: 25%
**发布后 (2024-10)**:
- 平均情感: -1.2
- 负面评论: 45%
**变化**: 情感显著恶化,负面评论增加 80%
# [主题] 情感分析报告
## 概况
- 分析时间: [时间范围]
- 数据量: X 条评论/讨论
- 数据来源: [列表]
## 整体情感分布
| 情感 | 数量 | 占比 |
|------|------|------|
| 正面 | X | XX% |
| 中性 | X | XX% |
| 负面 | X | XX% |
**平均情感评分**: X.X (-5 到 +5)
## 痛点严重程度排名
1. [痛点1] - 严重程度: XX/20
2. [痛点2] - 严重程度: XX/20
3. [痛点3] - 严重程度: XX/20
## 高频关键词
### 正面
1. [词1] - 提及 X 次
2. [词2] - 提及 X 次
### 负面
1. [词1] - 提及 X 次
2. [词2] - 提及 X 次
## 主题分布
| 主题 | 数量 | 占比 | 平均情感 |
|------|------|------|---------|
| [主题1] | X | XX% | X.X |
## 竞品对比
[竞品情感对比表]
## 时间趋势
[情感趋势图]
## 典型评论样本
### 最正面
> "[引用]"
### 最负面
> "[引用]"
## 洞察与建议
1. [基于情感分析的洞察]
2. [建议]
-- 情感分析结果表
CREATE TABLE sentiment_analysis (
id INTEGER PRIMARY KEY,
quote_id INTEGER,
sentiment TEXT, -- positive/neutral/negative
sentiment_score REAL, -- -5 to +5
intensity TEXT, -- low/medium/high
theme TEXT,
keywords TEXT, -- JSON array
analyzed_at TEXT,
FOREIGN KEY (quote_id) REFERENCES user_quotes(id)
);
-- 关键词频率表
CREATE TABLE keyword_frequency (
id INTEGER PRIMARY KEY,
keyword TEXT,
category TEXT,
frequency INTEGER,
sentiment_association TEXT, -- positive/negative
last_updated TEXT
);