Skip to content
forked from kiwirafe/xiangsi

中文文本相似度计算器 + 中文文本聚类

License

Notifications You must be signed in to change notification settings

hamutama/xiangshi

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

50 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

相识(Xiangshi)

中文文本相似度计算器 + 中文聚类计算器

Chinese Text Similarity Calculator + Text Clustering (English Version)

Pypi Version Pypi Downloads Pypi and Github License Language

相识的优势有:

  • 专攻中文文本相似度比较
  • 支持余弦、Simhash、Minhash和Ngram算法
  • 100%Python语言
  • 可以单独计算TF和IDF
  • 安装容易
  • 100%开源
  • 长期维护和更新

下载与安装

Pip安装:

pip3 install xiangshi

国内较慢的话可以使用清华镜像:

pip3 install -i https://pypi.tuna.tsinghua.edu.cn/simple xiangshi

新版本

v4.0.0:

  • 再次减少20%时间
  • 增加Kmeans聚类
  • 重新修改Cache方法
  • 重新修改Input输入

v4.1.0:

  • 修改Kmeans中variation的计算
  • 输出格式修改

注意:

  • v4.0.0的计算结果可能和v3.0.0不一样因为v4.0.0不再自动TFIDF加权
  • 注意:v4.0.0中的Input均为list

使用方法

计算文本相似度

余弦相似度示例

import xiangshi as xs
xs.cossim(["我曾经失落失望失掉所有方向", "直到看见平凡才是唯一的答案"])

Simhash & Minhash & Ngram相似度示例

import xiangshi as xs
# Simhash
xs.simhash(["我曾经失落失望失掉所有方向", "直到看见平凡才是唯一的答案"])
# Minhash
xs.minhash(["我曾经失落失望失掉所有方向", "直到看见平凡才是唯一的答案"])
# Ngram(适用于英文文本)
xs.ngram(["When life gives you lemons", "eat watermelons"])

聚类

计算Kmeans Clustering示例

import xiangshi as xs
arg = [
    "我曾经失落失望失掉所有方向", 
    "直到看见平凡才是唯一的答案", 
    "我曾经跨过山和大海", 
    "也穿过人山人海", 
    "转眼都飘散如烟",
    "我曾经拥有着的一切", 
    "也哭也笑平凡着",
    "那片笑声让我想起我的那些花儿",
]
# 第一个输入值为组的数量,需要聚类成三个组就为3
# 第二个输入值为Data
# 第三个输入值为是否需要Clusters的中心,True为需要,False为不需要默认值为False
xs.kmeans(3, arg, WithKeys=True) 

注意:kmeans自动TFIDF加权且用欧几里得距离算出文本之间的距离

计算Kmeans的K值示例

import xiangshi as xs
arg = [
    "我曾经失落失望失掉所有方向", 
    "直到看见平凡才是唯一的答案", 
    "我曾经跨过山和大海", 
    "也穿过人山人海", 
    "转眼都飘散如烟",
    "我曾经拥有着的一切", 
    "也哭也笑平凡着",
    "那片笑声让我想起我的那些花儿",
]
# 第一个输入值为Data
xs.calk(arg) 

注意:calk()从0到MaxNum算出每个Kmeans并找出最优K值,其中因为算量较大所以会导致时间较长。下一个版本将会支持以C为内核的多线程Kmeans

计算TF,IDF,TFIDF

import xiangshi as xs
xs.GetTF(corpus) # corpus为文本,必须先分词好
xs.GetIDF(corpus, lists) # corpus和lists为文本,必须先分词好
xs.GetTFIDF(corpus, lists) # corpus和lists为文本,不用分词

其它加权方法

xs.weight = "tf"
xs.weight = "tfidf"

如果是TFIDF加权的话将用于IDF的文本加在原来的list后面。 假如原来是: `xs.cossim(["我曾经失落失望失掉所有方向", "直到看见平凡才是唯一的答案"])` 则TFIDF加权为

arg = [
    "我曾经失落失望失掉所有方向", 
    "直到看见平凡才是唯一的答案", 
    "我曾经跨过山和大海", # 用于TFIDF
    "也穿过人山人海", # 用于TFIDF
    "转眼都飘散如烟", # 用于TFIDF
    "我曾经拥有着的一切", # 用于TFIDF
    "也哭也笑平凡着", # 用于TFIDF
    "那片笑声让我想起我的那些花儿", # 用于TFIDF
]
xs.cossim(arg)

其他函数

import xiangshi as xs
xs.file2list(file1, file2, EncodeArg="utf-8") # v3.0.0用文件输入的可以先用这个转为list再进行文本相似度计算
xs.SegDepart(corpus) # 分词+停用词过滤
xs.dict2file(dict) # Dict到File
xs.SortDict(dict) # Dict排序

修改默认函数

import xiangshi as xs
#计算TFIDF时是否使用log,True是使用,False是不使用。默认值为True
xs.UseLog = True
#计算Simhash时取前多少的TFIDF值。默认值为64
xs.feature = 64
#计算Minhash时算出多少个哈希值。默认值为16
xs.HashNums = 16
#计算Minhash时的最大哈希。默认值为4294967311
xs.prime = 4294967311

其他链接

相识寓意

同是天涯沦落人,相逢何必曾相识

MIT License

Copyright (c) [2021] [Kiwirafe]

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

About

中文文本相似度计算器 + 中文文本聚类

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%