【448】NLP, NER, PoS

目录：

1. 停用词 —— stopwords
2. 介词 —— prepositions —— part of speech
3. Named Entity Recognition (NER)　　3.1 Stanford NER
   　　3.2 spaCy
   　　3.3 NLTK
4. 句子中单词提取（Word extraction）

---

1. 停用词（stopwords）

ref: Removing stop words with NLTK in Python

ref: Remove Stop Words

import nltk
# nltk.download('stopwords')
from nltk.corpus import stopwords
print(stopwords.words('english'))

output:
['i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', "you're", "you've", "you'll", "you'd", 'your', 'yours', 'yourself', 'yourselves', 'he', 'him', 'his', 'himself', 'she', "she's", 'her', 'hers', 'herself', 'it', "it's", 'its', 'itself', 'they', 'them', 'their', 'theirs', 'themselves', 'what', 'which', 'who', 'whom', 'this', 'that', "that'll", 'these', 'those', 'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being', 'have', 'has', 'had', 'having', 'do', 'does', 'did', 'doing', 'a', 'an', 'the', 'and', 'but', 'if', 'or', 'because', 'as', 'until', 'while', 'of', 'at', 'by', 'for', 'with', 'about', 'against', 'between', 'into', 'through', 'during', 'before', 'after', 'above', 'below', 'to', 'from', 'up', 'down', 'in', 'out', 'on', 'off', 'over', 'under', 'again', 'further', 'then', 'once', 'here', 'there', 'when', 'where', 'why', 'how', 'all', 'any', 'both', 'each', 'few', 'more', 'most', 'other', 'some', 'such', 'no', 'nor', 'not', 'only', 'own', 'same', 'so', 'than', 'too', 'very', 's', 't', 'can', 'will', 'just', 'don', "don't", 'should', "should've", 'now', 'd', 'll', 'm', 'o', 're', 've', 'y', 'ain', 'aren', "aren't", 'couldn', "couldn't", 'didn', "didn't", 'doesn', "doesn't", 'hadn', "hadn't", 'hasn', "hasn't", 'haven', "haven't", 'isn', "isn't", 'ma', 'mightn', "mightn't", 'mustn', "mustn't", 'needn', "needn't", 'shan', "shan't", 'shouldn', "shouldn't", 'wasn', "wasn't", 'weren', "weren't", 'won', "won't", 'wouldn', "wouldn't"]

2. 介词（prepositions, part of speech）

ref: How do I remove verbs, prepositions, conjunctions etc from my text? [closed]

ref: Alphabetical list of part-of-speech tags used in the Penn Treebank Project:

>>> import nltk
>>> sentence = """At eight o'clock on Thursday morning
... Arthur didn't feel very good."""
>>> tokens = nltk.word_tokenize(sentence)
>>> tokens
['At', 'eight', "o'clock", 'on', 'Thursday', 'morning',
'Arthur', 'did', "n't", 'feel', 'very', 'good', '.']
>>> tagged = nltk.pos_tag(tokens)
>>> tagged[0:6]
[('At', 'IN'), ('eight', 'CD'), ("o'clock", 'JJ'), ('on', 'IN'),
('Thursday', 'NNP'), ('morning', 'NN')]

3. Named Entity Recognition (NER)

ref: Introduction to Named Entity Recognition

ref: Named Entity Recognition with NLTK and SpaCy

• Standford NER
• spaCy
• NLTK

3.1 Stanford NER

article = '''
Asian shares skidded on Tuesday after a rout in tech stocks put Wall Street to the sword, while a
sharp drop in oil prices and political risks in Europe pushed the dollar to 16-month highs as investors dumped
riskier assets. MSCI’s broadest index of Asia-Pacific shares outside Japan dropped 1.7 percent to a 1-1/2
week trough, with Australian shares sinking 1.6 percent. Japan’s Nikkei dived 3.1 percent led by losses in
electric machinery makers and suppliers of Apple’s iphone parts. Sterling fell to $1.286 after three straight
sessions of losses took it to the lowest since Nov.1 as there were still considerable unresolved issues with the
European Union over Brexit, British Prime Minister Theresa May said on Monday.'''

import nltk
from nltk.tag import StanfordNERTagger

print('NTLK Version: %s' % nltk.__version__)

stanford_ner_tagger = StanfordNERTagger(
    r"D:\Twitter Data\Data\NER\stanford-ner-2018-10-16\classifiers\english.muc.7class.distsim.crf.ser.gz",
	r"D:\Twitter Data\Data\NER\stanford-ner-2018-10-16\stanford-ner-3.9.2.jar"
)

results = stanford_ner_tagger.tag(article.split())

print('Original Sentence: %s' % (article))
for result in results:
    tag_value = result[0]
    tag_type = result[1]
    if tag_type != 'O':
        print('Type: %s, Value: %s' % (tag_type, tag_value))

output:
NTLK Version: 3.4
Original Sentence:
Asian shares skidded on Tuesday after a rout in tech stocks put Wall Street to the sword, while a
sharp drop in oil prices and political risks in Europe pushed the dollar to 16-month highs as investors dumped
riskier assets. MSCI’s broadest index of Asia-Pacific shares outside Japan dropped 1.7 percent to a 1-1/2
week trough, with Australian shares sinking 1.6 percent. Japan’s Nikkei dived 3.1 percent led by losses in
electric machinery makers and suppliers of Apple’s iphone parts. Sterling fell to $1.286 after three straight
sessions of losses took it to the lowest since Nov.1 as there were still considerable unresolved issues with the
European Union over Brexit, British Prime Minister Theresa May said on Monday.
Type: DATE, Value: Tuesday
Type: LOCATION, Value: Europe
Type: ORGANIZATION, Value: Asia-Pacific
Type: LOCATION, Value: Japan
Type: PERCENT, Value: 1.7
Type: PERCENT, Value: percent
Type: ORGANIZATION, Value: Nikkei
Type: PERCENT, Value: 3.1
Type: PERCENT, Value: percent
Type: LOCATION, Value: European
Type: LOCATION, Value: Union
Type: PERSON, Value: Theresa
Type: PERSON, Value: May

3.2 spaCy

import spacy
from spacy import displacy
from collections import Counter
import en_core_web_sm
nlp = en_core_web_sm.load()
doc = nlp(article)
for X in doc.ents:
	print('Value: %s, Type: %s' % (X.text, X.label_))

output:
Value: Asian, Type: NORP
Value: Tuesday, Type: DATE
Value: Europe, Type: LOC
Value: MSCI’s, Type: ORG
Value: Asia-Pacific, Type: LOC
Value: Japan, Type: GPE
Value: 1.7 percent, Type: PERCENT
Value: 1-1/2, Type: CARDINAL
Value: Australian, Type: NORP
Value: 1.6 percent, Type: PERCENT
Value: Japan, Type: GPE
Value: 3.1 percent, Type: PERCENT
Value: Apple, Type: ORG
Value: 1.286, Type: MONEY
Value: three, Type: CARDINAL
Value: Nov.1, Type: NORP
Value: the
European Union, Type: ORG
Value: Brexit, Type: GPE
Value: British, Type: NORP
Value: Theresa May, Type: PERSON
Value: Monday, Type: DATE

标签含义：https://spacy.io/api/annotation#pos-tagging

Type	Description
PERSON	People, including fictional.
NORP	Nationalities or religious or political groups.
FAC	Buildings, airports, highways, bridges, etc.
ORG	Companies, agencies, institutions, etc.
GPE	Countries, cities, states.
LOC	Non-GPE locations, mountain ranges, bodies of water.
PRODUCT	Objects, vehicles, foods, etc. (Not services.)
EVENT	Named hurricanes, battles, wars, sports events, etc.
WORK_OF_ART	Titles of books, songs, etc.
LAW	Named documents made into laws.
LANGUAGE	Any named language.
DATE	Absolute or relative dates or periods.
TIME	Times smaller than a day.
PERCENT	Percentage, including ”%“.
MONEY	Monetary values, including unit.
QUANTITY	Measurements, as of weight or distance.
ORDINAL	“first”, “second”, etc.
CARDINAL	Numerals that do not fall under another type.

3.3 NLTK

import nltk
from nltk import word_tokenize, pos_tag, ne_chunk
nltk.download('words')
nltk.download('averaged_perceptron_tagger')
nltk.download('punkt')
nltk.download('maxent_ne_chunker')

def fn_preprocess(art):
    art = nltk.word_tokenize(art)
    art = nltk.pos_tag(art)
    return art
art_processed = fn_preprocess(article)
print(art_processed)

output:
[('Asian', 'JJ'), ('shares', 'NNS'), ('skidded', 'VBN'), ('on', 'IN'), ('Tuesday', 'NNP'), ('after', 'IN'), ('a', 'DT'), ('rout', 'NN'), ('in', 'IN'), ('tech', 'JJ'), ('stocks', 'NNS'), ('put', 'VBD'), ('Wall', 'NNP'), ('Street', 'NNP'), ('to', 'TO'), ('the', 'DT'), ('sword', 'NN'), (',', ','), ('while', 'IN'), ('a', 'DT'), ('sharp', 'JJ'), ('drop', 'NN'), ('in', 'IN'), ('oil', 'NN'), ('prices', 'NNS'), ('and', 'CC'), ('political', 'JJ'), ('risks', 'NNS'), ('in', 'IN'), ('Europe', 'NNP'), ('pushed', 'VBD'), ('the', 'DT'), ('dollar', 'NN'), ('to', 'TO'), ('16-month', 'JJ'), ('highs', 'NNS'), ('as', 'IN'), ('investors', 'NNS'), ('dumped', 'VBD'), ('riskier', 'JJR'), ('assets', 'NNS'), ('.', '.'), ('MSCI', 'NNP'), ('’', 'NNP'), ('s', 'VBD'), ('broadest', 'JJS'), ('index', 'NN'), ('of', 'IN'), ('Asia-Pacific', 'NNP'), ('shares', 'NNS'), ('outside', 'IN'), ('Japan', 'NNP'), ('dropped', 'VBD'), ('1.7', 'CD'), ('percent', 'NN'), ('to', 'TO'), ('a', 'DT'), ('1-1/2', 'JJ'), ('week', 'NN'), ('trough', 'NN'), (',', ','), ('with', 'IN'), ('Australian', 'JJ'), ('shares', 'NNS'), ('sinking', 'VBG'), ('1.6', 'CD'), ('percent', 'NN'), ('.', '.'), ('Japan', 'NNP'), ('’', 'NNP'), ('s', 'VBD'), ('Nikkei', 'NNP'), ('dived', 'VBD'), ('3.1', 'CD'), ('percent', 'NN'), ('led', 'VBN'), ('by', 'IN'), ('losses', 'NNS'), ('in', 'IN'), ('electric', 'JJ'), ('machinery', 'NN'), ('makers', 'NNS'), ('and', 'CC'), ('suppliers', 'NNS'), ('of', 'IN'), ('Apple', 'NNP'), ('’', 'NNP'), ('s', 'VBD'), ('iphone', 'NN'), ('parts', 'NNS'), ('.', '.'), ('Sterling', 'NN'), ('fell', 'VBD'), ('to', 'TO'), ('$', '$'), ('1.286', 'CD'), ('after', 'IN'), ('three', 'CD'), ('straight', 'JJ'), ('sessions', 'NNS'), ('of', 'IN'), ('losses', 'NNS'), ('took', 'VBD'), ('it', 'PRP'), ('to', 'TO'), ('the', 'DT'), ('lowest', 'JJS'), ('since', 'IN'), ('Nov.1', 'NNP'), ('as', 'IN'), ('there', 'EX'), ('were', 'VBD'), ('still', 'RB'), ('considerable', 'JJ'), ('unresolved', 'JJ'), ('issues', 'NNS'), ('with', 'IN'), ('the', 'DT'), ('European', 'NNP'), ('Union', 'NNP'), ('over', 'IN'), ('Brexit', 'NNP'), (',', ','), ('British', 'NNP'), ('Prime', 'NNP'), ('Minister', 'NNP'), ('Theresa', 'NNP'), ('May', 'NNP'), ('said', 'VBD'), ('on', 'IN'), ('Monday', 'NNP'), ('.', '.')]

　　

4. 句子中单词提取（Word extraction）

ref: An introduction to Bag of Words and how to code it in Python for NLP

import re
def word_extraction(sentence):
	ignore = ['a', "the", "is"]
	words = re.sub("[^\w]", " ",  sentence).split()
	cleaned_text = [w.lower() for w in words if w not in ignore]
	return cleaned_text

a = "alex is. good guy."
print(word_extraction(a))

output:
['alex', 'good', 'guy']