From f7fe44c1528d97b9f9936d25380e28329c2c6c7e Mon Sep 17 00:00:00 2001 From: wangzhuc Date: Mon, 30 Mar 2026 23:23:51 +0800 Subject: [PATCH] fix: expand negative markers and vocabulary temperature word lists MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit NEGATIVE_MARKERS: 26 → 51 words Added: despair (绝望/迷茫/心累), deception (骗/忽悠/割韭菜/套路), failure (白费/黄了/凉了), self-deprecation (傻/天真/自嗨), sarcasm (呵呵/行吧/真服了), complaint (受够了/苦哈哈) COLD_WORDS: 7 → 25 (技术栈/标准化/护城河/飞轮/底层逻辑/PMF/ROI...) WARM_WORDS: 7 → 15 (老实说/这么说吧/你想啊/有意思的是...) HOT_WORDS: 8 → 19 (凡尔赛/标题党/躺平/摆烂/破防/上头/内耗...) WILD_WORDS: 7 → 17 (苦哈哈/傻乎乎/交学费/踩坑/翻车...) Impact on 15 exemplar articles: neg score avg: 0.15 → 0.27 (+80%) temp_mix: still low on short segments, but full articles now score 0.33-1.00 vs previously 0.00 Co-Authored-By: Claude Opus 4.6 (1M context) --- scripts/humanness_score.py | 38 ++++++++++++++++++++++++++++++++++---- 1 file changed, 34 insertions(+), 4 deletions(-) diff --git a/scripts/humanness_score.py b/scripts/humanness_score.py index 6aabc32..ea7e0e0 100644 --- a/scripts/humanness_score.py +++ b/scripts/humanness_score.py @@ -55,10 +55,23 @@ REAL_SOURCE_PATTERNS = [ ] NEGATIVE_MARKERS = [ + # 直接负面情绪 "失望", "糟糕", "扯", "坑", "烂", "差劲", "崩溃", "吐槽", "骂", "怒", "烦", "焦虑", "担忧", "不满", "恶心", "可怕", "可悲", "可笑", "离谱", "尴尬", "无语", "蠢", "惨", "亏", "危", + # 绝望/迷茫 + "绝望", "迷茫", "心累", "丧", "后悔", "后怕", "心寒", + # 欺骗/操控（隐性负面） + "骗", "忽悠", "割韭菜", "套路", "画大饼", "洗脑", + # 失败/徒劳 + "白费", "白搭", "没戏", "黄了", "凉了", "废了", + # 自嘲/自贬 + "傻", "天真", "吃亏", "自嗨", "打脸", + # 讽刺/反语 + "呵呵", "好吧", "行吧", "真服了", + # 短语 "太扯了", "说实话我很失望", "搞什么", "不靠谱", "受不了", + "受够了", "想哭", "伤心", "苦哈哈", "得过且过", ] COMMON_ADVERBS = [ @@ -69,10 +82,27 @@ COMMON_ADVERBS = [ "竟然", "简直", "几乎", "完全", "绝对", "必然", ] -COLD_WORDS = ["边际", "认知负荷", "信息不对称", "路径依赖", "商业模式", "生态系统", "增量"] -WARM_WORDS = ["说白了", "其实吧", "讲真", "说实话", "坦白讲", "懂的都懂", "怎么说呢"] -HOT_WORDS = ["DNA动了", "格局打开", "遥遥��先", "卷", "内卷", "炸了", "杀疯了", "吃灰"] -WILD_WORDS = ["整挺好", "不靠谱", "瞎折腾", "搁这儿", "糊弄", "扯", "嗯"] +COLD_WORDS = [ + "边际", "认知负荷", "信息不对称", "路径依赖", "商业模式", "生态系统", "增量", + "技术栈", "标准化", "结构性", "规模化", "护城河", "飞轮", "闭环", + "赛道", "壁垒", "方法论", "底层逻辑", "第一性原理", "杠杆", "复利", + "ROI", "PMF", "代运营", "供给侧", "需求侧", +] +WARM_WORDS = [ + "说白了", "其实吧", "讲真", "说实话", "坦白讲", "懂的都懂", "怎么说呢", + "老实说", "这么说吧", "你想啊", "别急", "慢慢来", + "有意思的是", "好玩的是", "巧的是", "说来话长", "话说回来", +] +HOT_WORDS = [ + "DNA动了", "格局打开", "遥遥领先", "卷", "内卷", "炸了", "杀疯了", "吃灰", + "凡尔赛", "标题党", "躺平", "摆烂", "破防", "上头", "内耗", + "蒸发", "出圈", "降维打击", "弯道超车", +] +WILD_WORDS = [ + "整挺好", "不靠谱", "瞎折腾", "搁这儿", "糊弄", "扯", "嗯", + "苦哈哈", "傻乎乎", "稀里糊涂", "得了吧", "算了吧", + "摔了跤", "交学费", "踩坑", "翻车", "栽了", +] SELF_CORRECTION_PATTERNS = [ r'不对[，,]', r'准确说', r'算了', r'说错了',