计算给定单词与随机单词列表之间的相似度。

Question

我想计算出一个给定的单词和一个随机的单词列表之间的相似性，然后在一个新的列表中对结果进行排序，例如：

list = ['bark','black','cat','bite','human','book'] #it could be another list

与以下词相似：

word = ['dog']

--

import spacy
nlp = spacy.load('en_core_web_md')


bark = nlp("bark")
bite = nlp("bite")
human = nlp("human")
book = nlp("book")
cat = nlp("cat")
black = nlp("black")

print("dog - bark", dog.similarity(bark)) #0.4258176903285793
print("dog - bite", dog.similarity(bite)) #0.4781574605069981
print("dog - human", dog.similarity(human)) #0.35814872466230835
print("dog - book", dog.similarity(book)) #0.22838638167627964
print("dog - cat", dog.similarity(cat)) #0.8016854705531046
print("dog - black", dog.similarity(black)) #0.30601667459001575

那么，我如何能够自动计算列表中每个单词与给定单词的相似度呢？

Answer 1

你可以这样做：

import spacy
nlp = spacy.load('en_core_web_md')

words = ['bark','black','cat','bite','human','book']
word = 'dog'
word_nlp = nlp(word)

new_words = [(w, word_nlp.similarity(nlp(w))) for w in words]
new_words.sort(key=lambda x: x[1], reverse=True)

for w, value in new_words:
    print(f"{word} - {w}", value)