从 gensim 解释负 Word2Vec 相似度

Question

例如我们使用 gensim:

训练一个 word2vec 模型

from gensim import corpora, models, similarities
from gensim.models.word2vec import Word2Vec

documents = ["Human machine interface for lab abc computer applications",
              "A survey of user opinion of computer system response time",
              "The EPS user interface management system",
              "System and human system engineering testing of EPS",
              "Relation of user perceived response time to error measurement",
              "The generation of random binary unordered trees",
              "The intersection graph of paths in trees",
              "Graph minors IV Widths of trees and well quasi ordering",
              "Graph minors A survey"]

texts = [[word for word in document.lower().split()] for document in documents]
w2v_model = Word2Vec(texts, size=500, window=5, min_count=1)

而当我们查询单词之间的相似度时，我们发现负相似度分数：

>>> w2v_model.similarity('graph', 'computer')
0.046929569156789336
>>> w2v_model.similarity('graph', 'system')
0.063683518562347399
>>> w2v_model.similarity('survey', 'generation')
-0.040026775040430063
>>> w2v_model.similarity('graph', 'trees')
-0.0072684112978664561

我们如何解读负分？

如果是余弦相似度，范围不应该是[0,1]吗？

Word2Vec.similarity(x,y)函数的上界和下界是多少？文档里写的不多：https://radimrehurek.com/gensim/models/word2vec.html#gensim.models.word2vec.Word2Vec.similarity =(

查看 Python 包装器代码，内容不多：https://github.com/RaRe-Technologies/gensim/blob/develop/gensim/models/word2vec.py#L1165

（如果可能，请给我指出实现相似度函数的 .pyx 代码。）

Answer 1

余弦相似度范围为-1到1，与正余弦波相同。

至于来源：

https://github.com/RaRe-Technologies/gensim/blob/ba1ce894a5192fc493a865c535202695bb3c0424/gensim/models/word2vec.py#L1511

def similarity(self, w1, w2):
    """
    Compute cosine similarity between two words.
    Example::
      >>> trained_model.similarity('woman', 'man')
      0.73723527
      >>> trained_model.similarity('woman', 'woman')
      1.0
    """
    return dot(matutils.unitvec(self[w1]), matutils.unitvec(self[w2])

Answer 2

正如其他人所说，根据所比较的两个向量之间的角度，余弦相似度的范围可以从 -1 到 1。 gensim 中的确切实现是归一化向量的简单点积。

https://github.com/RaRe-Technologies/gensim/blob/4f0e2ae0531d67cee8d3e06636e82298cb554b04/gensim/models/keyedvectors.py#L581

def similarity(self, w1, w2):
        """
        Compute cosine similarity between two words.
        Example::
          >>> trained_model.similarity('woman', 'man')
          0.73723527
          >>> trained_model.similarity('woman', 'woman')
          1.0
        """
        return dot(matutils.unitvec(self[w1]), matutils.unitvec(self[w2]))

在解释方面，您可以像考虑相关系数一样考虑这些值。值为1表示词向量之间存在完美关系（如"woman"与"woman"相比），值为0表示词之间没有关系，值为-1表示词向量之间存在完全相反的关系单词。

从 gensim 解释负 Word2Vec 相似度

Interpreting negative Word2Vec similarity from gensim

python

nlp

similarity

gensim

word2vec