├── stop_words
├── README.md
├── words_mining.py
└── words.hongloumeng


/stop_words:
--------------------------------------------------------------------------------
  1 | 的
  2 | 了
  3 | 着
  4 | 这
  5 | 在
  6 | 是
  7 | 去
  8 | 却
  9 | 让
 10 | 你
 11 | 我
 12 | 他
 13 | 她
 14 | 它
 15 | 们
 16 | 令人
 17 | 已被
 18 | 得很
 19 | 结果
 20 | 已毕
 21 | 并没
 22 | 性情
 23 | 家务
 24 | 原故
 25 | 正值
 26 | 现场
 27 | 功劳
 28 | 实在
 29 | 为何
 30 | 全部
 31 | 全都
 32 | 只有
 33 | 所有
 34 | 事情
 35 | 出现
 36 | 身体
 37 | 身躯
 38 | 身材
 39 | 还请
 40 | 身份
 41 | 体积
 42 | 体型
 43 | 身形
 44 | 听完
 45 | 之后
 46 | 之物
 47 | 就是
 48 | 还是
 49 | 还有
 50 | 有点
 51 | 其他
 52 | 都没
 53 | 没有
 54 | 还请
 55 | 承受
 56 | 担心
 57 | 达到
 58 | 模样
 59 | 照片
 60 | 背景
 61 | 关系
 62 | 并不
 63 | 无法
 64 | 情节
 65 | 接著
 66 | 什麽
 67 | 一眼
 68 | 一个
 69 | 一丝
 70 | 一些
 71 | 一定
 72 | 一旁
 73 | 一句
 74 | 一声
 75 | 一阵
 76 | 一径
 77 | 一伸
 78 | 一愣
 79 | 一笑
 80 | 一边
 81 | 效果
 82 | 这个
 83 | 已经
 84 | 之后
 85 | 之意
 86 | 之恩
 87 | 之罪
 88 | 大胆
 89 | 自己
 90 | 我们
 91 | 你们
 92 | 多久
 93 | 他们
 94 | 算是
 95 | 数量
 96 | 等人
 97 | 像是
 98 | 都是
 99 | 什么
100 | 也就
101 | 也是
102 | 也可
103 | 也会
104 | 随之
105 | 根本
106 | 从来
107 | 从此
108 | 最后
109 | 却是
110 | 极为
111 | 差点
112 | 感觉
113 | 名字
114 | 唤做
115 | 方才
116 | 双手
117 | 谢恩
118 | 即忙
119 | 即唤
120 | 即命
121 | 今已
122 | 证据
123 | 脑袋
124 | 远超
125 | 吓得
126 | 情况
127 | 威觉
128 | 交代
129 | 交流
130 | 问题
131 | 现场
132 | 好便似
133 | 好道也
134 | 也不敢
135 | 不弱于
136 | 不可能
137 | 这么多
138 | 能不能
139 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # 基于自由度及凝固度的新词发现/新词挖掘
 2 | 
 3 | ## 流程图
 4 | 
 5 | <div style="align: center">
 6 | <img src="https://github.com/elephantnose/FigureBed/blob/master/words_mining_flowe_graph2.PNG"/>
 7 | </div>
 8 | 
 9 | ## 四大名著样例
10 | 
11 |  TOP50(按词频降序排序)
12 | 
13 | 《红楼梦》 
14 | 宝玉, 凤姐, 贾母, 黛玉, 袭人, 姑娘, 王夫人, 宝钗, 丫头, 老太太, 贾政, 奶奶, 众人, 贾琏, 老爷, 东西, 姐姐, 大家, 二爷, 薛姨妈, 凤姐儿, 探春, 鸳鸯, 紫鹃, 湘云, 婆子, 妹妹, 贾珍, 银子, 李纨, 晴雯, 尤氏, 媳妇, 外头, 刘姥姥, 薛蟠, 邢夫人, 小丫头, 孩子, 林黛玉, 姊妹, 香菱, 麝月, 哥哥, 丫鬟, 贾蓉, 小厮, 意思, 二奶奶, 主意
15 | 
16 | 《西游记》 
17 | 行者, 八戒, 师父, 三藏, 大圣, 唐僧, 沙僧, 和尚, 菩萨, 长老, 妖精, 老孙, 悟空, 国王, 那怪, 徒弟, 闻言, 大王, 小妖, 兄弟, 宝贝, 孙行者, 铁棒, 龙王, 妖怪, 师徒, 太子, 东土, 性命, 孙大圣, 老爷, 那呆子, 神通, 公主, 妖魔, 人家, 玉帝, 猴王, 哥哥, 土地, 道士, 师兄, 贫僧, 行李, 云头, 陛下, 太宗, 那妖精, 闻得, 爷爷
18 | 
19 | 《水浒传》 
20 | 宋江, 李逵, 武松, 林冲, 军马, 哥哥, 吴用, 头领, 太尉, 众人, 戴宗, 兄弟, 卢俊义, 梁山泊, 先锋, 燕青, 好汉, 花荣, 王庆, 晁盖, 石秀, 那妇人, 杨志, 鲁智深, 柴进, 呼延灼, 太公, 山寨, 秦明, 和尚, 史进, 天子, 张顺, 公孙胜, 兄长, 弟兄, 关胜, 军士, 朱仝, 知府, 张清, 庄客, 杨雄, 李俊, 性命, 那厮, 小弟, 东京, 小喽罗, 大官人
21 | 
22 | 《三国演义》 
23 | 玄德, 孔明, 将军, 曹操, 司马, 二人, 丞相, 关公, 引兵, 云长, 荆州, 蜀兵, 夏侯, 张飞, 吕布, 诸葛, 主公, 魏延, 孙权, 赵云, 军士, 魏兵, 刘备, 司马懿, 夫人, 姜维, 袁绍, 东吴, 诸将, 周瑜, 汉中, 都督, 马超, 陛下, 天子, 后主, 黄忠, 张郃, 先主, 太守, 先生, 邓艾, 孟获, 先锋, 诸葛亮, 汝等, 江东, 曹仁, 张辽, 领兵
24 | 
25 | ## 计算速度
26 | 单进程 40K/s
27 | 
28 | 
29 | 
30 | 参考文章: [互联网时代的社会语言学：基于SNS的文本数据挖掘][1]
31 | 
32 | [1]: http://www.matrix67.com/blog/archives/5044
33 | 


--------------------------------------------------------------------------------
/words_mining.py:
--------------------------------------------------------------------------------
  1 | import os
  2 | import re
  3 | import time
  4 | import math
  5 | 
  6 | 
  7 | def loda_word_dict():
  8 |     """
  9 |     加载
 10 |     结巴词典词性: https://github.com/elephantnose/words_mining/blob/master/dict.txt.big
 11 |     停用词: https://github.com/elephantnose/words_mining/blob/master/stop_words
 12 |     *结巴自带词典词性标注准确度一般, 可根据需要修改
 13 |     """
 14 |     word_dict = {}
 15 |     stop_words = {}
 16 | 
 17 |     if os.path.exists("./dict.txt.big"):
 18 |         with open("./dict.txt.big", "r", encoding="utf8") as fr:
 19 |             word_dict = {word.strip().split(" ")[0]: word.strip().split(" ")[2] for word in fr}
 20 | 
 21 |     if os.path.exists("./stop_words"):
 22 |         with open("./stop_words", encoding="utf8") as fr:
 23 |             stop_words = set([word.strip() for word in fr])
 24 |     
 25 |     return word_dict, stop_words
 26 | 
 27 | 
 28 | class Config(object):
 29 |     __author__ = "elephantnose"
 30 |     __github__ = "https://github.com/elephantnose"
 31 |     __jianshu__ = "https://www.jianshu.com/u/39efcd8e2587"
 32 |     __segmentfault__ = "https://segmentfault.com/u/elephantnose"
 33 |     
 34 |     """
 35 |     各阈值设定
 36 |     words_length: 新词字数默认个字
 37 |     pmi_limit: 凝固度阈值
 38 |     left_entropy_limit: 左熵阈值
 39 |     right_entropy_limit: 右熵阈值
 40 |     word_frequency: 词频阈值
 41 |     """
 42 |     words_length = 5
 43 |     pmi_limit = 1.5
 44 |     left_entropy_limit = 1
 45 |     right_entropy_limit = 1
 46 |     word_frequency_limit = 5
 47 |     word_dict, stop_words = loda_word_dict()
 48 | 
 49 | 
 50 | class ContentHandler(object):
 51 |     def __init__(self, content):
 52 |         """
 53 |         length: 文本字数
 54 |         book_content: 正序文本
 55 |         r_book_content: 倒序文本
 56 |         tire_tree: 正序文本 tire树
 57 |         r_tire_tree: 倒序文本 tire树
 58 |         """
 59 |         self.length = 0
 60 |         self.book_content, self.r_book_content = self._preprocess(content=content)
 61 |         
 62 |         self.tire_tree = TireTree(self.book_content)
 63 |         self.r_tire_tree = TireTree(self.r_book_content)
 64 |     
 65 |     def pmi(self, char_node):
 66 |         """计算凝固度"""
 67 |         p_x_y = char_node.count / self.length
 68 |         px_py_list = []
 69 |         
 70 |         # 枚举所有组成词的情况, 并取最大概率值
 71 |         for i in range(1, len(char_node.name)):
 72 |             px = self.tire_tree.search_node(char_node.name[:i]).count / self.length
 73 |             py = self.tire_tree.search_node(char_node.name[i:]).count / self.length
 74 |             px_py_list.append(px*py)
 75 | 
 76 |         px_py = max(px_py_list)
 77 |         p = math.log10(p_x_y / px_py)
 78 |         return p
 79 |     
 80 |     def left_entropy(self, char_node):
 81 |         """计算左熵"""
 82 |         r_char_node = self.r_tire_tree.search_node(char_node.name[::-1])
 83 |         father_set = r_char_node.child
 84 |         le = 0
 85 |         for father_name, father_node in father_set.items():
 86 |             p_father = father_node.count / r_char_node.child_counts
 87 |             p = p_father * math.log10(p_father)
 88 |             le += p
 89 |         return -le
 90 |     
 91 |     def right_entropy(self, char_node):
 92 |         """计算右熵"""
 93 |         child_set = self.tire_tree.search_node(char_node.name).child
 94 |         re = 0
 95 |         for child_name, child_node in child_set.items():
 96 |             p_child = child_node.count / char_node.child_counts
 97 |             p = p_child * math.log10(p_child)
 98 |             re += p
 99 |         return -re
100 | 
101 |     def word_frequency(self, char_node):
102 |         """计算词频"""
103 |         return char_node.count
104 |     
105 |     def get_words(self, node, layer, res_data):
106 |         if layer >= self.tire_tree.layer-1:
107 |             return
108 | 
109 |         for c_name, c_node in node.child.items():
110 |             # 递归
111 |             self.get_words(c_node, layer+1, res_data)
112 | 
113 |             # 纯小写英文及纯数字过滤
114 |             if (c_name.encode("utf8").isalpha() and c_name.encode("utf8").islower()) or c_name.encode("utf8").isdigit():
115 |                 continue
116 |             # 词典过滤
117 |             is_continue = self.word_dict_filter(c_name)
118 |             if not is_continue:
119 |                 continue            
120 |             # 阈值过滤
121 |             plrf = self.limit_filter(c_node)
122 |             if not plrf:
123 |                 continue
124 | 
125 |             res_data.append(plrf)
126 | 
127 |     def limit_filter(self, node):
128 |         wf = node.count
129 |         if wf < Config.word_frequency_limit:
130 |             return False
131 | 
132 |         pmi = self.pmi(node)
133 |         if pmi < Config.pmi_limit:
134 |             return False
135 | 
136 |         le = self.left_entropy(node)
137 |         if le < Config.left_entropy_limit:
138 |             return False
139 | 
140 |         re = self.right_entropy(node)
141 |         if re < Config.right_entropy_limit:
142 |             return False
143 | 
144 |         return [node.name, wf, pmi, le, re]
145 | 
146 |     def word_dict_filter(self, chars):
147 |         """词典过滤"""
148 |         for char in self.permutation(1, chars):
149 |             # 过滤掉已经收录于结巴词典的非名词
150 |             if not Config.word_dict.get(char, "n").startswith("n"):
151 |                 return False
152 |             # 过滤停用词
153 |             if char in Config.stop_words:
154 |                 return False
155 |         return True
156 | 
157 |     def permutation(self, start_size, char):
158 |         """字符排列组合"""
159 |         for size in range(start_size, len(char)+1):
160 |             for index in range(len(char)+1-size):
161 |                 yield char[index: index+size]
162 | 
163 |     def _preprocess(self, content):
164 |         """返回正序文本及倒序文本列表, 按符号拆分"""
165 |         content_list = re.split("[^\u4E00-\u9FFFa-zA-Z0-9]", content)
166 |         r_content_list = re.split("[^\u4E00-\u9FFFa-zA-Z0-9]", content[::-1])
167 | 
168 |         if not self.length:
169 |             self.length = sum([len(i) for i in content_list if i])
170 | 
171 |         return content_list, r_content_list
172 | 
173 | 
174 | class Node(object):
175 |     def __init__(self, name, father):
176 |         """
177 |         节点名称
178 |         节点出现次数
179 |         父节点
180 |         子节点列表
181 |         未去重子集量
182 |         """
183 |         self.name = name
184 |         self.count = 0
185 |         self.father = father
186 |         self.child = {}
187 |         self.child_counts = 0
188 | 
189 | 
190 | class TireTree(object):
191 |     def __init__(self, 
192 |         content, 
193 |         layer_num=Config.words_length+1, 
194 |         step=1):
195 |         """
196 |         字典树对象
197 |         layer_num: 字典树层数
198 |         content: 构建字典树的字符串
199 |         """
200 |         self.content = content
201 |         self.root = Node("ROOT", None)
202 |         self.layer = layer_num
203 |         self.step = step
204 |         self.word_counts = 0
205 |         
206 |         self.build_tree()
207 | 
208 |     def build_tree(self):
209 |         # 按层构建字典树, layer 从1开始, 表示第一层, 第0层为根节点
210 |         for layer in range(1, self.layer+1):
211 |             # 创建切割窗口对象
212 |             char_session = CharSession(size=layer, step=self.step)
213 |             for char in char_session.split_char(self.content):
214 |                 # 判断 char 是否在该层 没有节点则添加, 有则更新
215 |                 if not self.search_node(char, layer):
216 |                     self.add_node(char, layer)
217 |                 else:
218 |                     self.update_node(char, layer)
219 |         
220 |     def add_node(self, char, layer=None):
221 |         """
222 |         在指定层添加指定字符串节点
223 |         """
224 |         if not layer:
225 |             layer = len(char)
226 | 
227 |         # 创建节点对象
228 |         if layer == 1:
229 |             father = self.root
230 |         else:
231 |             father = self.search_node(char[:-1])
232 | 
233 |         node = Node(name=char, father=father)
234 |         node.count = 1
235 | 
236 |         # 将此节点挂入tire树
237 |         father.child[char] = node
238 |         father.child_counts += 1
239 |         return True
240 | 
241 |     def del_node(self, char, layer=None):
242 |         pass
243 | 
244 |     def update_node(self, char, layer=None):
245 |         """
246 |         更新指定节点信息
247 |         """
248 |         if not layer:
249 |             layer = len(char)
250 | 
251 |         node = self.search_node(char, layer)
252 |         node.count += 1
253 |         node.father.child_counts += 1
254 |         return True
255 | 
256 |     def search_node(self, char, layer=None):
257 |         """
258 |         指定字符串, 指定层 查找节点是否存在
259 |         """
260 |         if char == "ROOT":
261 |             return self.root
262 |         elif not layer:
263 |             layer = len(char)
264 | 
265 |         node = self.root
266 |         for layer_index in range(1, layer+1):
267 |             node = node.child.get(char[:layer_index], None)
268 |             if not node:
269 |                 return None
270 |         
271 |         return node
272 | 
273 | 
274 | class CharSession(object):
275 |     def __init__(self, size, step=1):
276 |         """
277 |         窗口对象
278 |         size: 窗口大小
279 |         step: 移动步长
280 |         """
281 |         self.size = size
282 |         self.step = step
283 | 
284 |     def split_char(self, content):
285 |         """
286 |         按指定窗口大小及步长切割文本
287 |         """
288 |         for seq in content:
289 |             while seq:
290 |                 if len(seq) >= self.size:
291 |                     yield seq[:self.size]
292 |                     seq = seq[self.step:]
293 |                 else:
294 |                     break
295 | 
296 | 
297 | def find_word(file_like):
298 |     """
299 |     file_like: 文本内容或文本路径
300 |     """
301 |     if os.path.exists(file_like):
302 |         with open(file_like, encoding="utf8") as fr:
303 |             content = fr.read()
304 |     else:
305 |         content = file_like
306 | 
307 |     content_handler = ContentHandler(content)
308 |     words = []
309 | 
310 |     for child_node in content_handler.tire_tree.search_node("ROOT").child.values():
311 |         content_handler.get_words(child_node, layer=1, res_data=words)
312 | 
313 |     return words
314 | 
315 | 
316 | if __name__ == '__main__':
317 |     stime = time.time()
318 | 
319 |     res_data = find_word("./hongloumeng.txt")
320 | 
321 |     for each_ele in res_data:
322 |         print(*each_ele, sep="\t")
323 | 
324 |     etime = time.time()
325 |     print("ALL DONE! 耗时 {} s".format(etime-stime))
326 | 


--------------------------------------------------------------------------------
/words.hongloumeng:
--------------------------------------------------------------------------------
  1 | 宝玉	3861	1.920249695660435	1.8683176015164489	2.008151912411824
  2 | 凤姐	1696	2.275124596786582	1.7668487030119284	1.6380616776396568
  3 | 贾母	1623	2.031328880152934	1.6402792484677713	1.8720408641173383
  4 | 黛玉	1317	2.082141604276614	1.3953585334549905	1.91362302465702
  5 | 袭人	1129	1.8210406779736985	1.7331325814552954	1.825454318223042
  6 | 姑娘	1097	2.496967933291215	1.7031773172478153	1.9506689103882342
  7 | 王夫人	1065	1.8402287433548559	1.5802223696419404	1.8642857385020823
  8 | 宝钗	1048	2.0871233075074116	1.7386438236061736	1.8692063744580105
  9 | 丫头	969	2.23731541908231	1.3194072121984664	1.844804920230693
 10 | 老太太	946	2.1492575135373446	1.7774458681510545	1.9395640257048075
 11 | 贾政	912	2.132523384311854	1.6999817333454295	1.843800906243123
 12 | 奶奶	846	2.2839769647542743	1.3761953348430698	2.047871938223925
 13 | 众人	834	1.703348563597313	1.5659121110474876	1.5729242846707672
 14 | 贾琏	764	2.085585127936571	1.7552048027535454	1.8841105963529399
 15 | 老爷	638	2.006942917138157	1.7615066285152836	2.000439762118316
 16 | 东西	584	2.843978170464574	1.4587457769845338	1.8539709720758677
 17 | 姐姐	579	1.5600462902044658	1.5996945301764403	1.9537517096113668
 18 | 二爷	496	1.897262354148805	1.478486490822326	2.019948791132446
 19 | 薛姨妈	453	2.8027356654731035	1.5713701725230598	1.7643545136821832
 20 | 凤姐儿	451	1.5029221997147726	1.604483628877994	1.7051544369753253
 21 | 探春	428	2.8027749677486096	1.6245619336969461	1.657029991585864
 22 | 鸳鸯	426	3.209579987836052	1.6268124409154276	1.7681143376631647
 23 | 紫鹃	416	3.147255291244207	1.5307701614640774	1.8131400985222335
 24 | 湘云	394	2.8394749131649806	1.045556330403844	1.6507707586826375
 25 | 婆子	392	1.9469532701585468	1.081621517364775	1.7302313100945823
 26 | 妹妹	390	2.3215042503787076	1.287503631044515	1.9107197787171628
 27 | 贾珍	384	2.008835906825046	1.6716602143105526	1.8137807198175355
 28 | 银子	362	1.9153151722484163	1.3218141118396858	1.6883526875269577
 29 | 李纨	359	3.019340005686991	1.5298303229786305	1.597757693637426
 30 | 晴雯	338	3.2984114846165857	1.646512294315143	1.7571766995585902
 31 | 尤氏	338	2.978673419736106	1.6235918474430888	1.7023080352427158
 32 | 媳妇	317	3.192794387643852	1.498207203046196	1.8527035621945067
 33 | 外头	314	1.7533887708167075	1.3222011233131437	1.882111856100834
 34 | 刘姥姥	293	3.0455788425703787	1.4059942841206015	1.6304486651592753
 35 | 薛蟠	284	2.837619363552421	1.685418814980996	1.8505170187878108
 36 | 邢夫人	283	1.8402287433548559	1.667419299459424	1.7291522687552563
 37 | 小丫头	275	2.0147226659882445	1.2464278857974624	1.3452231538009827
 38 | 孩子	250	2.0101141525141037	1.0219335790144668	1.6251820373882222
 39 | 林黛玉	249	2.084443814418511	1.367402241373067	1.7558237514585642
 40 | 姊妹	247	2.752092861141172	1.5301659534875365	1.474738042407255
 41 | 香菱	241	2.8887437740650905	1.6512189824986296	1.8790629552762699
 42 | 麝月	233	2.932531915435284	1.4324506205217118	1.512570954939784
 43 | 哥哥	230	2.5250966915000497	1.3421007156264415	1.8502219499810209
 44 | 丫鬟	229	2.7666080150789183	1.4214167951894632	1.5941761442487457
 45 | 贾蓉	224	2.0153347892181097	1.475562172457836	1.8060627943594254
 46 | 小厮	212	2.4971145030341195	1.2617759440290894	1.4631055762142229
 47 | 二奶奶	209	1.838338516824433	1.4969280832987364	1.8207428710118478
 48 | 主意	201	2.35357723398317	1.1231725941765343	1.4167625700981725
 49 | 惜春	195	2.700535797254927	1.5027015629896956	1.6377491552823815
 50 | 贾芸	191	2.0560817187703218	1.2809803285833437	1.685368108137415
 51 | 贾赦	188	2.0937386003176433	1.6996524707924239	1.5666510102884166
 52 | 雨村	183	3.2469244950313993	1.2682676993439737	1.7969732300400767
 53 | 嫂子	180	2.059775106871739	1.320929711722148	1.7250800110874873
 54 | 兄弟	180	3.0540585069713693	1.2787164734532022	1.8760567372531505
 55 | 母亲	174	1.7406039675379534	1.0613764177689347	1.8324175714368378
 56 | 素日	173	2.1121415853864356	1.4646164340126455	1.8863904967574552
 57 | 芳官	158	2.856166987360585	1.540783425612464	1.723180213673603
 58 | 金桂	153	2.891757665760179	1.5369577011967788	1.5796429029927048
 59 | 妙玉	152	1.849646895025003	1.3888967513586365	1.6530102617813554
 60 | 贾环	151	1.9467260600742642	1.5250966645371395	1.556825062712983
 61 | 雪雁	147	3.353379210064482	1.3124134562881018	1.6721377095428114
 62 | 迎春	142	2.545180585217271	1.5874329334451118	1.7021202724264213
 63 | 奴才	136	2.4465478220558246	1.3293526877206518	1.6157211773376368
 64 | 赵姨娘	132	2.5576267401990345	1.4225518242127881	1.7094382677092332
 65 | 衣服	132	2.8501725784733782	1.3420341604416457	1.6541386143739198
 66 | 林姑娘	132	2.0483536457842275	1.5302460913640068	1.7332787874185303
 67 | 莺儿	130	2.036778820599282	1.5045717658027198	1.6218718165925505
 68 | 府里	130	1.5202859688851877	1.268319759623021	1.5206564009430266
 69 | 和尚	119	2.2688787855663004	1.1629743299053878	1.561667906450683
 70 | 宝蟾	117	2.085586271776365	1.4858028950409359	1.6228263060831636
 71 | 宝琴	117	1.891909837919571	1.49128394831757	1.6324843146737815
 72 | 妈妈	116	1.980441737517973	1.3271901757696414	1.6741004472224266
 73 | 亲戚	110	2.7578076882803284	1.3675028229825108	1.3533020389655708
 74 | 地方	108	2.013091659803695	1.1847732720361819	1.372555727613529
 75 | 秦钟	107	3.26586151431019	1.4643613601434027	1.708315093074342
 76 | 姨太太	107	1.6361861101509727	1.493844288873282	1.6424906261420298
 77 | 光景	105	3.139127931698933	1.003122699259533	1.2570217505692511
 78 | 老婆子	104	1.8696514267891124	1.2067355014299854	1.5611997189892064
 79 | 薛蝌	102	2.8476022729596813	1.3710046675713963	1.6029214798522224
 80 | 父母	100	1.973901353507686	1.6133192428399632	1.6861954824038432
 81 | 秋纹	99	3.2413020971732776	1.0628765744289488	1.500343103494676
 82 | 衣裳	96	3.200650747571753	1.3910367184470929	1.4022682063551253
 83 | 林妹妹	94	2.3500495887262884	1.3437158107723286	1.551174469853685
 84 | 岫烟	94	3.4644669064647613	1.1721462863970828	1.6179077531491255
 85 | 贾兰	91	1.9083446598593719	1.2074451710530374	1.5846531980656087
 86 | 尤二姐	91	2.314782157370063	1.256882502758549	1.6075655580689003
 87 | 小丫头子	91	1.6464305949931832	1.1671278327377754	1.455560157892341
 88 | 婶子	89	1.8857802049769736	1.4263217150512204	1.6442323657891802
 89 | 赖大	88	1.9720535841114635	1.3481817823021764	1.3570497166214297
 90 | 茗烟	88	3.2094253476481542	1.2936676187805067	1.462697289790773
 91 | 工夫	87	2.3355545161254727	1.1700450977613046	1.472683057527292
 92 | 先生	86	1.782602162880173	1.2490782999849188	1.6725413374666984
 93 | 史湘云	83	2.936615468150483	1.1167266692561717	1.4818584627411038
 94 | 书房	81	1.9447220420871507	1.1980029450630048	1.1122918056832731
 95 | 侄儿	79	1.8552246658420986	1.4724532966944543	1.6062397302877012
 96 | 老祖宗	78	2.277300916847658	1.4434654963717144	1.5071576787489693
 97 | 荣府	74	2.543049146462673	1.4175367779887604	1.326531569472486
 98 | 士隐	74	3.2989707821593113	1.0974329396418676	1.5351919473584572
 99 | 贾瑞	73	1.5298405861108517	1.2202277529836547	1.4791824959295714
100 | 秦氏	73	2.6793532470755475	1.3496486086534079	1.5503521948873955
101 | 旺儿	72	1.9487322679225307	1.3360715182633534	1.4139028374643523
102 | 叔叔	72	2.9811882373757586	1.3368973539377464	1.5939565763911745
103 | 眼睛	70	3.001888507252217	1.2518272343083605	1.447296889754225
104 | 吃茶	70	1.8560314821799047	1.5123749522725076	1.1836914970589403
105 | 女孩儿	68	1.7835756200989357	1.1345292598453884	1.244650590369866
106 | 司棋	68	3.4821914891145633	1.3376474758949746	1.5279235635487676
107 | 读书	67	2.8118313631458163	1.6056695829482026	1.2338558116407277
108 | 规矩	67	3.9007439146222627	1.2841283941227133	1.4106706871416514
109 | 闻得	66	1.7453538415426273	1.3481817823021767	1.648778844169048
110 | 太监	66	1.9744595331288537	1.1442856792286127	1.469948286599035
111 | 贾蔷	63	1.965029670906714	1.209981208498597	1.40668475109739
112 | 眼泪	59	2.2484165327398427	1.1969913744053495	1.288687099355286
113 | 宝兄弟	59	1.6157652868854915	1.2744462746869218	1.568213117869548
114 | 琥珀	58	4.087651707165842	1.0845840056546714	1.3612742599196843
115 | 焙茗	58	3.646959718072854	1.2019459763759048	1.216485355672154
116 | 彩云	57	2.4163427585876653	1.2183584300502075	1.4258106636882955
117 | 包勇	56	3.451838759036323	1.1293529119916217	1.400320382450168
118 | 晚饭	54	2.3438612018060754	1.0205192698167445	1.1236701124308353
119 | 宁府	54	2.5767293876986415	1.3423372781277623	1.3357335131551762
120 | 倒象	53	1.8851311291385087	1.001773499021995	1.2508068690320109
121 | 并无	51	1.6331655459032257	1.2570217505692516	1.5281985296598823
122 | 大观园	51	2.3264239883710016	1.3059165694524901	1.1339577373391931
123 | 鲍二	50	2.420256728487632	1.2879122273284427	1.2252133255197561
124 | 风流	50	2.581543407972812	1.5792881853887293	1.3892526180087998
125 | 舅舅	50	2.8456846688420727	1.2390210148604943	1.4691843482137972
126 | 巧姐儿	50	1.7728201363108833	1.2602934735623719	1.4862307299936792
127 | 代儒	50	3.7104459755939607	1.056942025200422	1.3119144044890476
128 | 文章	49	3.282877976661784	1.3166251430300524	1.1680505452985441
129 | 北静王	47	2.594454060611855	1.2209650647939894	1.2943079330492295
130 | 元妃	46	3.2726149012144616	1.2672962814527096	1.4908927082394159
131 | 亲友	46	2.421576683356765	1.4457130351072818	1.3290657973597024
132 | 海棠	45	3.523380276727279	1.3230654384702145	1.2537106030277168
133 | 扇子	45	1.8454359590341283	1.274604822671226	1.29548525189965
134 | 师父	44	2.944304024159872	1.1659601640196429	1.5031606305853589
135 | 弟兄	43	2.43225445744765	1.1600794788105424	1.3729390241768458
136 | 邢王二夫人	42	1.8402287433548559	1.184420824864272	1.3028034946859763
137 | 玉钏儿	42	1.799208085937762	1.1424355007077336	1.3432307444612757
138 | 混帐	42	2.8384433058454066	1.1275442038508285	1.1051638171517637
139 | 打谅	42	2.5188914049100717	1.2028877460153864	1.1657608993919972
140 | 丰儿	42	1.8689118834491711	1.1162337185377762	1.3286609289269573
141 | 角门	41	2.3174068232255873	1.2926055547460202	1.1659315219598279
142 | 厅上	41	1.917000226687207	1.0811621533994946	1.3170731724059996
143 | 作诗	40	1.9144629260642272	1.4130175530116131	1.048843810057643
144 | 针线	39	3.5891766167264594	1.2281339676616938	1.1692558083232352
145 | 金钏儿	39	1.9616659366860698	1.1781710659629392	1.4257431431580816
146 | 张华	39	2.831963410281706	1.2125707966370325	1.3851623617739424
147 | 倪二	39	2.364529342843184	1.0244373361363204	1.352145250701002
148 | 颜色	38	3.05392154019883	1.175714509968042	1.2991925804077487
149 | 菩萨	38	4.238295844009043	1.3343515271072905	1.3205710848716485
150 | 板儿	38	1.703712592934992	1.0516015891793142	1.329185343066291
151 | 大嫂子	38	1.6509350798845057	1.0977958056755082	1.3405575876193938
152 | 金陵	37	2.9376238522247884	1.1062373268101808	1.2310779570619612
153 | 神仙	37	2.503428007265168	1.1217611155679588	1.2287224654717086
154 | 李嬷嬷	37	2.4357453431927816	1.0132944582093153	1.234789286917152
155 | 银钱	36	1.8906327077713856	1.1547757200762878	1.2122438289850557
156 | 金荣	36	2.1648531959863924	1.195261601815712	1.296312483049863
157 | 精神	36	2.669807929169317	1.3221618768308905	1.3922799373649233
158 | 甄宝玉	36	1.741501153824392	1.1083941944520206	1.3851623617739424
159 | 王仁	36	2.2915720600862803	1.2122438289850554	1.3701266324953756
160 | 月钱	36	1.7675412494986393	1.0266378908108658	1.1561994984127493
161 | 族中	36	2.316250089389043	1.37256000258626	1.1351571219224417
162 | 坠儿	36	1.886285979518594	1.2107592620876937	1.288928464604505
163 | 二哥哥	36	1.6401372585020968	1.091137509132795	1.258556136010284
164 | 二叔	36	1.6382531146275454	1.0493381639457804	1.2848996986361956
165 | 秋桐	35	3.2795372334047546	1.15231566512091	1.3592466188729728
166 | 新鲜	35	3.074067380381816	1.2812285866959998	1.213572880038171
167 | 尤三姐	35	2.3119892571858927	1.0857961488469916	1.2768882211571222
168 | 荣国府	34	2.985053120873014	1.1593349918922735	1.167630178225665
169 | 照管	34	1.941901373626415	1.34075177440518	1.2226279650578922
170 | 梅花	34	2.5370265002774586	1.3715192480087866	1.2593485078862006
171 | 园门	34	1.5394236759916733	1.0494004530231424	1.100982673971512
172 | 藕官	33	2.7780838316120122	1.0886293931650515	1.3114547368203622
173 | 甄家	33	1.7106287979422312	1.2222606640310345	1.1609416736221438
174 | 琏二爷	33	1.7286433789846491	1.2609309280316907	1.2815181461465335
175 | 春燕	33	2.4371706142299274	1.1031237505833833	1.2194636332345474
176 | 女婿	33	2.817948497952052	1.2935651950091311	1.3100293267182603
177 | 饮食	32	3.3301051623032345	1.1681337652831403	1.0886293931650515
178 | 婆婆	32	1.8101202923201596	1.1819257557628253	1.3014011874747542
179 | 箱子	31	1.7855746776303887	1.25270812939223	1.0425513192597389
180 | 桂花	31	2.048946251091027	1.296312483049863	1.2055908646113058
181 | 庙里	31	1.6376759197526574	1.22114767275494	1.2885271841708332
182 | 山石	31	2.5973978483122715	1.1016302863790104	1.0142697749229166
183 | 蒋玉菡	29	4.307011656378503	1.1214813959193146	1.3497752093658009
184 | 荷包	29	3.1810258858987317	1.1196192149346	1.15231566512091
185 | 琏二奶奶	29	2.0478628273848614	1.0527487960856206	1.322348733920363
186 | 珍大爷	29	2.270355233080358	1.1405288753313518	1.2374070497365839
187 | 比先	29	1.6552182045734707	1.2471415861560746	1.085958301447765
188 | 桃花	29	2.515587215569386	1.2603235910956663	1.1104159301996999
189 | 张罗	29	2.8297620577239355	1.4001159298305461	1.060119465488529
190 | 帖儿	29	1.5960868315062946	1.0046303971855928	1.0627843095674208
191 | 夫妻	29	2.0160411147674786	1.1686120220698188	1.185898936242587
192 | 题目	28	2.84156810264724	1.0634105123887099	1.026637890810866
193 | 男女	28	2.11942187750539	1.3100293267182606	1.2935497713373494
194 | 宁国府	28	2.9976422481810343	1.1242030405556924	1.012450248147257
195 | 众姊妹	28	1.8474945484160505	1.0791812460476247	1.2306771003829997
196 | 芙蓉	27	3.352411849936793	1.2703994189276515	1.1204119982655922
197 | 舅母	27	1.6033109526104048	1.0441456077951246	1.2453135859556532
198 | 益发	27	2.3925817436295778	1.0405296753565512	1.2891562186230305
199 | 玻璃	27	4.373958446009117	1.0703221291572096	1.3300395757676093
200 | 手帕子	27	1.9898839365943446	1.1647752718815907	1.044981826173945
201 | 女尼	27	2.605859585624852	1.0746090020803172	1.0232249084065224
202 | 李氏	26	1.7492797703980105	1.0958165935452864	1.2110849003808126
203 | 房屋	26	1.560163689861027	1.0757790619213747	1.1359539263004839
204 | 冯紫英	26	4.063292361306397	1.0601194654885289	1.1063788149324716
205 | 雀儿	25	1.7126835208666074	1.02599452062265	1.2603235910956663
206 | 铁槛寺	25	4.065749865718011	1.0703221291572096	1.0880336186233166
207 | 贵妃	25	2.726784196062831	1.1951419200116562	1.0556720202883954
208 | 胭脂	25	4.045986822386106	1.0961844950135782	1.1788716777510708
209 | 盒子	24	1.6436102780935022	1.0293335830649735	1.044981826173945
210 | 珍珠	24	2.3375735550102417	1.0405296753565512	1.1596183341632196
211 | 王子腾	24	2.6383595463109364	1.108356562103204	1.1050717496266387
212 | 形容	24	2.861592803490504	1.2876899543378464	1.2499943283000725
213 | 含泪	24	2.665348722183645	1.1596183341632194	1.1260332722757374
214 | 分派	24	2.082672865518848	1.2055908646113058	1.1757522959833626
215 | 金银	23	1.6331325273240023	1.0962320868634188	1.1680505452985441
216 | 胡乱	23	2.4923174682963203	1.0438877668097386	1.158731857656642
217 | 省亲	23	1.874865329069792	1.21537886502357	1.0179571597331107
218 | 彩霞	23	3.1714148515881466	1.0003923426013537	1.1596183341632196
219 | 家伙	23	1.7579626767378616	1.1947885434319039	1.0676310452816091
220 | 阴阳	22	3.1813825457464717	1.06608149544718	1.1549291732153124
221 | 贾赦等	22	1.644724913196067	1.1031237505833835	1.2603235910956663
222 | 素习	22	2.8581447914018097	1.098458818336966	1.3150563175800265
223 | 柳湘莲	22	2.987162323770918	1.0912337342819318	1.1288624837399295
224 | 俱已	22	1.9877865878244654	1.2041199826559248	1.2876899543378464
225 | 风月	21	1.6317322314219185	1.1664912331979271	1.1088567413913666
226 | 梨香院	21	3.3145212581572485	1.1107114985054047	1.048843810057643
227 | 祖父	20	2.245855486144543	1.1950336277707467	1.1692558083232354
228 | 情愿	20	2.0175947230999833	1.0394022484799927	1.0958165935452864
229 | 舌头	19	1.7335031626915223	1.0013342591828385	1.0438877668097386
230 | 笔砚	19	3.4907454305978702	1.1203167611296814	1.0003923426013537
231 | 碟子	19	1.814410890428682	1.0210614281209909	1.1950336277707467
232 | 册子	18	1.6577185320048693	1.0754089640953688	1.021318738256381
233 | 世兄	17	2.0120085212718264	1.041392685158225	1.1664912331979271
234 | 骨肉	16	3.2501068050420305	1.0601194654885289	1.041392685158225
235 | 顺路	16	2.555485114643579	1.021318738256381	1.0770309979379904
236 | 梯己	15	2.6857364352063198	1.0405296753565512	1.041392685158225
237 | 官员	13	2.577633226696886	1.021318738256381	1.041392685158225
238 | 


--------------------------------------------------------------------------------