Table of Contents
cntext is a text analysis package that provides traditional text analysis methods, such as word count, readability, document similarity, sentiment analysis, etc. It has built-in multiple Chinese and English sentiment dictionaries. Supporting word embedding models training and usage, cntext provides semantic distance and semantic projection now.
-
github repo
https://github.com/hidadeng/cntext -
pypi link
https://pypi.org/project/cntext/
By the day of 2023-11-20, the cumulative download volume of cntext reached 36581
pip install cntext --upgrade
pip install numpy==1.24.2
pip install gensim==4.2.0
pip install scikit-learn==1.1.2
import cntext as ct
help(ct)
Run
Help on package cntext:
NAME
cntext
PACKAGE CONTENTS
bias
dictionary
similarity
stats
Currently, the built-in functions of stats.py are:
- readability() the readability of text, support Chinese and English
- term_freq() word count
- dict_pkl_list() get the list of built-in dictionaries (pkl format) in cntext
- load_pkl_dict() load the pkl dictionary file
- sentiment() sentiment analysis
- sentiment_by_valence() valence sentiment analysis
import cntext as ct
text = 'What a sunny day!'
diction = {'Pos': ['sunny', 'good'],
'Neg': ['bad', 'terrible'],
'Adv': ['very']}
ct.sentiment(text=text,
diction=diction,
lang='english')
Run
{'Pos_num': 1,
'Neg_num': 0,
'Adv_num': 0,
'stopword_num': 1,
'word_num': 5,
'sentence_num': 1}
The larger the indicator, the higher the complexity of the article and the worse the readability.
readability(text, lang='chinese')
- text: text string
- lang: "chinese" or "english"，default is "chinese"
import cntext as ct
text = 'Committed to publishing quality research software with zero article processing charges or subscription fees.'
ct.readability(text=text,
lang='english')
Run
{'readability': 19.982}
Word count statistics function, return Counter type.
import cntext as ct
text = 'Committed to publishing quality research software with zero article processing charges or subscription fees.'
ct.term_freq(text=text, lang='english')
Run
Counter({'committed': 1,
'publishing': 1,
'quality': 1,
'research': 1,
'software': 1,
'zero': 1,
'article': 1,
'processing': 1,
'charges': 1,
'subscription': 1,
'fees.': 1})
get the list of built-in dictionaries (pkl format) in cntext
import cntext as ct
ct.dict_pkl_list()
Run
['DUTIR.pkl',
'HOWNET.pkl',
'sentiws.pkl',
'Chinese_Digitalization.pkl',
'ChineseFinancialFormalUnformalSentiment.pkl',
'Concreteness.pkl',
'ANEW.pkl',
'LSD2015.pkl',
'NRC.pkl',
'geninqposneg.pkl',
'HuLiu.pkl',
'AFINN.pkl',
'ChineseEmoBank.pkl',
'ADV_CONJ.pkl',
'Loughran_McDonald_Financial_Sentiment.pkl',
'Chinese_Loughran_McDonald_Financial_Sentiment.pkl',
'STOPWORDS.pkl']
We list 12 pkl dictionary here, some of English dictionary listed below are organized from quanteda.sentiment
| pkl文件 | 词典 | 语言 | 功能 |
|---|---|---|---|
| ChineseEmoBank.pkl | Chinese Sentiment Dictionary, includes 「valence」「arousal」. In cntext, we only take Chinese valence-arousal words (CVAW, single word) into account, ignore CVAP, CVAS, CVAT. | Chinese | valence, arousal |
| DUTIR.pkl | DUTIR | Chinese | Seven categories of emotions: 哀, 好, 惊, 惧, 乐, 怒, 恶 |
| HOWNET.pkl | Hownet | Chinese | Positive、Negative |
| SentiWS.pkl | SentimentWortschatz (SentiWS) | German | Positive、Negative； |
| ChineseFinancialFormalUnformalSentiment.pkl | Chinese finance dictionary, contains formal、unformal、positive、negative | Chinese | formal-pos、 formal-neg； unformal-pos、 unformal-neg |
| ANEW.pkl | Affective Norms for English Words (ANEW) | English | |
| LSD2015.pkl | Lexicoder Sentiment Dictionary (2015) | English | Positive、Negative |
| NRC.pkl | NRC Word-Emotion Association Lexicon | English | fine-grained sentiment words; |
| HuLiu.pkl | Hu&Liu (2004) | English | Positive、Negative |
| AFINN.pkl | Affective Norms for English Words | English | |
| ADV_CONJ.pkl | adverbial & conjunction | Chinese | |
| STOPWORDS.pkl | English&Chinese | stopwordlist | |
| Concreteness.pkl | Brysbaert, M., Warriner, A. B., & Kuperman, V. (2014). Concreteness ratings for 40 thousand generally known English word lemmas. Behavior Research Methods, 46, 904–911 | English | word & concreateness score |
| Chinese_Loughran_McDonald_Financial_Sentiment.pkl | 曾庆生, 周波, 张程, and 陈信元. "年报语调与内部人交易: 表里如一还是口是心非?." 管理世界 34, no. 09 (2018): 143-160. | Chinese | 正面、负面词 |
| Chinese_Digitalization.pkl | 吴非,胡慧芷,林慧妍,任晓怡. 企业数字化转型与资本市场表现——来自股票流动性的经验证据[J]. 管理世界,2021,37(07):130-144+10. | 中文 | 基于这篇论文，构建了中文数字化词典，含人工智能技术、大数据技术、云计算技术、区块链技术、数字技术应用等关键词列表。 |
| Loughran_McDonald_Financial_Sentiment.pkl | Loughran, Tim, and Bill McDonald. "When is a liability not a liability? Textual analysis, dictionaries, and 10‐Ks." The Journal of finance 66, no. 1 (2011): 35-65. | English | Positive and Negative emotion words in the financial field。 Besides, in version of 2018, author add Uncertainty, Litigious, StrongModal, WeakModal, Constraining |
| Chinese_FLS.pkl | 许帅,邵帅,何贤杰.业绩说明会前瞻性信息对分析师盈余预测准确性的影响——信口雌黄还是言而有征[J].中国管理科学:1-15. | 中文 | 前瞻性词典集，含174个词语 |
load the pkl dictionary file and return dict type data.
import cntext as ct
print(ct.__version__)
# load the pkl dictionary file
print(ct.load_pkl_dict('NRC.pkl'))
Run
1.8.0
{'NRC': {'anger': ['abandoned', 'abandonment', 'abhor', 'abhorrent', ...],
'anticipation': ['accompaniment','achievement','acquiring', ...],
'disgust': ['abject', 'abortion', 'abundance', 'abuse', ...],
'fear': ['anxiety', 'anxious', 'apache', 'appalling', ...],
......
'Desc': 'NRC Word-Emotion Association Lexicon',
'Referer': 'Mohammad, Saif M., and Peter D. Turney. "Nrc emotion lexicon." National Research Council, Canada 2 (2013).'
}
sentiment(text, diction, lang='chinese')
Calculate the occurrences of each emotional category words in text; The complex influence of adverbs and negative words on emotion is not considered.
- text: text string
- diction: emotion dictionary data, support diy or built-in dicitonary
- lang: "chinese" or "english"，default is "chinese"
We can use built-in dicitonary in cntext, such as NRC.pkl
import cntext as ct
text = 'What a happy day!'
ct.sentiment(text=text,
diction=ct.load_pkl_dict('NRC.pkl')['NRC'],
lang='english')
Run
{'anger_num': 0,
'anticipation_num': 1,
'disgust_num': 0,
'fear_num': 0,
'joy_num': 1,
'negative_num': 0,
'positive_num': 1,
'sadness_num': 0,
'surprise_num': 0,
'trust_num': 1,
'stopword_num': 1,
'word_num': 5,
'sentence_num': 1}
We can also use DIY dicitonary, just like
import cntext as ct
text = 'What a happy day!'
diction = {'Pos': ['happy', 'good'],
'Neg': ['bad', 'terrible'],
'Adv': ['very']}
ct.sentiment(text=text,
diction=diction,
lang='english')
Run
{'Pos_num': 1,
'Neg_num': 0,
'Adv_num': 0,
'stopword_num': 1,
'word_num': 5,
'sentence_num': 1}
sentiment_by_valence(text, diction, lang='english')
Calculate the occurrences of each sentiment category words in text; The complex influence of intensity adverbs and negative words on emotion is not considered.
- text: text sring
- diction: sentiment dictionary with valence.；
- lang: "chinese" or "english"; default language="english"
Here we want to study the concreteness of text. The concreteness.pkl that comes from Brysbaert2014.
Brysbaert, M., Warriner, A. B., & Kuperman, V. (2014). Concreteness ratings for 40 thousand generally known English word lemmas. Behavior Research Methods, 46, 904–911
import cntext as ct
# load the concreteness.pkl dictionary file; cntext version >=1.7.1
concreteness_df = ct.load_pkl_dict('concreteness.pkl')['concreteness']
concreteness_df.head()
Run
| word | valence | |
|---|---|---|
| 0 | roadsweeper | 4.85 |
| 1 | traindriver | 4.54 |
| 2 | tush | 4.45 |
| 3 | hairdress | 3.93 |
| 4 | pharmaceutics | 3.77 |
reply = "I'll go look for that"
score=ct.sentiment_by_valence(text=reply,
diction=concreteness_df,
lang='english')
score
Run
1.85
employee_replys = ["I'll go look for that",
"I'll go search for that",
"I'll go search for that top",
"I'll go search for that t-shirt",
"I'll go look for that t-shirt in grey",
"I'll go search for that t-shirt in grey"]
for idx, reply in enumerate(employee_replys):
score=ct.sentiment_by_valence(text=reply,
diction=concreteness_df,
lang='english')
template = "Concreteness Score: {score:.2f} | Example-{idx}: {exmaple}"
print(template.format(score=score,
idx=idx,
exmaple=reply))
ct.sentiment_by_valence(text=text, diction=concreteness_df, lang='english')
Run
Concreteness Score: 1.55 | Example-0: I'll go look for that
Concreteness Score: 1.55 | Example-1: I'll go search for that
Concreteness Score: 1.89 | Example-2: I'll go search for that top
Concreteness Score: 2.04 | Example-3: I'll go search for that t-shirt
Concreteness Score: 2.37 | Example-4: I'll go look for that t-shirt in grey
Concreteness Score: 2.37 | Example-5: I'll go search for that t-shirt in grey
This module is used to build or expand the vocabulary (dictionary), including
- SoPmi Co-occurrence algorithm to extend vocabulary (dictionary), Only support chinese
- W2VModels using word2vec to extend vocabulary (dictionary), support english & chinese
import cntext as ct
import os
sopmier = ct.SoPmi(cwd=os.getcwd(),
#raw corpus data，txt file.only support chinese data now.
input_txt_file='data/sopmi_corpus.txt',
#muanually selected seed words
seedword_txt_file='data/sopmi_seed_words.txt', #人工标注的初始种子词
)
sopmier.sopmi()
Run
Step 1/4:...Preprocess Corpus ...
Step 2/4:...Collect co-occurrency information ...
Step 3/4:...Calculate mutual information ...
Step 4/4:...Save candidate words ...
Finish! used 44.49 s
In particular, note that the code needs to set the lang parameter
import cntext as ct
import os
#init W2VModels, corpus data w2v_corpus.txt
model = ct.W2VModels(cwd=os.getcwd(), lang='english')
model.train(input_txt_file='data/w2v_corpus.txt')
#According to the seed word, filter out the top 100 words that are most similar to each category words
model.find(seedword_txt_file='data/w2v_seeds/integrity.txt',
topn=100)
model.find(seedword_txt_file='data/w2v_seeds/innovation.txt',
topn=100)
model.find(seedword_txt_file='data/w2v_seeds/quality.txt',
topn=100)
model.find(seedword_txt_file='data/w2v_seeds/respect.txt',
topn=100)
model.find(seedword_txt_file='data/w2v_seeds/teamwork.txt',
topn=100)
Run
Step 1/4:...Preprocess corpus ...
Step 2/4:...Train word2vec model
used 174 s
Step 3/4:...Prepare similar candidates for each seed word in the word2vec model...
Step 4/4 Finish! Used 187 s
Step 3/4:...Prepare similar candidates for each seed word in the word2vec model...
Step 4/4 Finish! Used 187 s
Step 3/4:...Prepare similar candidates for each seed word in the word2vec model...
Step 4/4 Finish! Used 187 s
Step 3/4:...Prepare similar candidates for each seed word in the word2vec model...
Step 4/4 Finish! Used 187 s
Step 3/4:...Prepare similar candidates for each seed word in the word2vec model...
Step 4/4 Finish! Used 187 s
When runing out the W2VModels, there will appear a file called w2v.model in the directory of output/w2v_candi_words.Note this w2v file can be used later.
from gensim.models import KeyedVectors
w2v_model = KeyedVectors.load("the path of w2v.model")
#to extract vector for word
#w2v_model.get_vector(word)
#if you need more information about the usage of w2_model, please use help function
#help(w2_model)
For example, we load the output/w2v_candi_words/w2v.model
from gensim.models import KeyedVectors
w2v_model = KeyedVectors.load('output/w2v_candi_words/w2v.model')
# find the most similar word in w2v.model
w2v_model.most_similar('innovation')
Run
[('technology', 0.689210832118988),
('infrastructure', 0.669672966003418),
('resources', 0.6695448160171509),
('talent', 0.6627111434936523),
('execution', 0.6549549102783203),
('marketing', 0.6533523797988892),
('merchandising', 0.6504817008972168),
('diversification', 0.6479553580284119),
('expertise', 0.6446896195411682),
('digital', 0.6326863765716553)]
#to extract vector for "innovation"
w2v_model.get_vector('innovation')
Run
array([-0.45616838, -0.7799563 , 0.56367606, -0.8570078 , 0.600359 ,
-0.6588043 , 0.31116748, -0.11956959, -0.47599426, 0.21840936,
-0.02268819, 0.1832016 , 0.24452794, 0.01084935, -1.4213187 ,
0.22840202, 0.46387577, 1.198386 , -0.621511 , -0.51598716,
0.13352732, 0.04140598, -0.23470387, 0.6402956 , 0.20394802,
0.10799981, 0.24908689, -1.0117126 , -2.3168423 , -0.0402851 ,
1.6886286 , 0.5357047 , 0.22932841, -0.6094084 , 0.4515793 ,
-0.5900931 , 1.8684244 , -0.21056202, 0.29313338, -0.221067 ,
-0.9535679 , 0.07325 , -0.15823542, 1.1477109 , 0.6716076 ,
-1.0096023 , 0.10605699, 1.4148282 , 0.24576302, 0.5740349 ,
0.19984631, 0.53964925, 0.41962907, 0.41497853, -1.0322098 ,
0.01090925, 0.54345983, 0.806317 , 0.31737605, -0.7965337 ,
0.9282971 , -0.8775608 , -0.26852605, -0.06743863, 0.42815775,
-0.11774074, -0.17956367, 0.88813037, -0.46279573, -1.0841943 ,
-0.06798118, 0.4493006 , 0.71962464, -0.02876493, 1.0282255 ,
-1.1993176 , -0.38734904, -0.15875885, -0.81085825, -0.07678922,
-0.16753489, 0.14065655, -1.8609751 , 0.03587054, 1.2792674 ,
1.2732009 , -0.74120265, -0.98000383, 0.4521185 , -0.26387128,
0.37045383, 0.3680011 , 0.7197629 , -0.3570571 , 0.8016917 ,
0.39243212, -0.5027844 , -1.2106236 , 0.6412354 , -0.878307 ],
dtype=float32)
generate word co-occurrence matrix
import cntext as ct
documents = ["I go to school every day by bus .",
"i go to theatre every night by bus"]
ct.co_occurrence_matrix(documents,
window_size=2,
lang='english')
Build the Glove model for english corpus data. corpus file path is data/brown_corpus.txt
import cntext as ct
import os
model = ct.Glove(cwd=os.getcwd(), lang='english')
model.create_vocab(file='data/brown_corpus.txt', min_count=5)
model.cooccurrence_matrix()
model.train_embeddings(vector_size=50, max_iter=25)
model.save()
Run
Step 1/4: ...Create vocabulary for Glove.
Step 2/4: ...Create cooccurrence matrix.
Step 3/4: ...Train glove embeddings.
Note, this part takes a long time to run
Step 3/4: ... Finish! Use 175.98 s
The generate生成的词嵌入模型文件位于output/Glove内
Four text similarity functions
- cosine_sim(text1, text2)
- jaccard_sim(text1, text2)
- minedit_sim(text1, text2)
- simple_sim(text1, text2)
Algorithm implementation reference from Cohen, Lauren, Christopher Malloy, and Quoc Nguyen. Lazy prices. No. w25084. National Bureau of Economic Research, 2018.
import cntext as ct
text1 = 'Programming is fun!'
text2 = 'Programming is interesting!'
print(ct.cosine_sim(text1, text2))
print(ct.jaccard_sim(text1, text2))
print(ct.minedit_sim(text1, text2))
print(ct.simple_sim(text1, text2))
Run
0.67
0.50
1.00
0.90
Word embeddings contain human cognitive information.
- tm.sematic_distance(words, c_words1, c_words2)
- tm.sematic_projection(words, c_words1, c_words2)
Calculate the two semantic distance， and return the difference between the two.
- words concept words, words = ['program', 'software', 'computer']
- c_words1 concept words1, c_words1 = ["man", "he", "him"]
- c_words2 concept words2, c_words2 = ["woman", "she", "her"]
For example,
male_concept = ['male', 'man', 'he', 'him']
female_concept = ['female', 'woman', 'she', 'her']
software_engineer_concept = ['engineer', 'programming', 'software']
d1 = distance(male_concept, software_engineer_concept)
d2 = distance(female_concept, software_engineer_concept)
If d1-d2<0，it means in semantic space, between man and woman, software_engineer_concept is more closer to male_concept。
In other words, there is a stereotype (bias) of women for software engineers in this corpus.
download glove_w2v.6B.100d.txt from google Driver
import cntext as ct
#Note: this is a word2vec format model
tm = ct.Text2Mind(w2v_model_path='glove_w2v.6B.100d.txt')
engineer = ['program', 'software', 'computer']
mans = ["man", "he", "him"]
womans = ["woman", "she", "her"]
tm.sematic_distance(words=animals,
c_words1=mans,
c_words2=womans)
Run
-0.38
-0.38 means in semantic space, engineer is closer to man, other than woman.
To explain the semantic projection of the word vector model, I use the picture from a Nature paper in 2022[@Grand2022SemanticPR]. Regarding the names of animals, human cognition information about animal size is hidden in the corpus text. By projecting the meaning of LARGE WORDS and SMALL WORDS with the vectors of different animals, the projection of the animal on the size vector(just like the red line in the bellow picture) is obtained, so the size of the animal can be compared by calculation.
Calculate the projected length of each word vector in the concept vector.Note that the calculation result reflects the direction of concept.Greater than 0 means semantically closer to c_words2.
Grand, G., Blank, I.A., Pereira, F. and Fedorenko, E., 2022. Semantic projection recovers rich human knowledge of multiple object features from word embeddings. Nature Human Behaviour, pp.1-13.
For example, in the corpus, perhaps show that our human beings have different size memory(perception) about animals.
animals = ['mouse', 'cat', 'horse', 'pig', 'whale']
small_words = ["small", "little", "tiny"]
large_words = ["large", "big", "huge"]
tm.sematic_projection(words=animals,
c_words1=small_words,
c_words2=large_words)
Run
[('mouse', -1.68),
('cat', -0.92),
('pig', -0.46),
('whale', -0.24),
('horse', 0.4)]
Regarding the perception of size, humans have implied in the text that mice are smaller and horses are larger.
If you use cntext in your research or in your project, please cite:
Deng X., Nan P. (2022). cntext: a Python tool for text mining (version 1.7.9). DOI: 10.5281/zenodo.7063523 URL: https://github.com/hiDaDeng/cntext
@misc{YourReferenceHere,
author = {Deng, Xudong and Nan, Peng},
doi = {10.5281/zenodo.7063523},
month = {9},
title = {cntext: a Python tool for text mining},
url = {https://github.com/hiDaDeng/cntext},
year = {2022}
}
%0 Generic
%A Deng, Xudong
%A Nan, Peng
%D 2022
%K text mining
%K text analysi
%K social science
%K management science
%K semantic analysis
%R 10.5281/zenodo.7063523
%T cntext: a Python tool for text mining
%U https://github.com/hiDaDeng/cntext