利用余弦相似性、TFIDF和pysp匹配Python中的公司名称

Question

我试图将两个列表中的公司名称进行匹配，以检查列表a中的某个公司是否确实列在B列表中。由于a公司的名称以各种不同的形式书写，我倾向于使用余弦相似性进行匹配。 为此，我关注了Ran Tavory在博客上的留言：datasets-of-company-1be4b1b2871e" rel="nofollow noreferrer">Link Here

以下是概要：

1. Calculate TF-IDF matrices on the driver.
2. Parallelize matrix A; Broadcast matrix B
3. Each worker now flatMaps its chunk of work by multiplying its chunk of matrix A with the entire matrix B. So if a worker operates on A[0:99] then it would multiply these hundred rows and return the result of, say A[13] matches a name found in B[21]. Multiplication is done using numpy.
4. The driver would collect back all the results from the different workers and match the indices (A[13] and B[21]) to the actual names in the original dataset — and we’re done!

我可以运行注释中描述的代码，但其中有一部分似乎有点奇怪： b_mat_dist = broadcast_matrix(a_mat)

当广播一个_mat以及并行化一个_mat时，我会得到一个逻辑结果，即每个公司的名称都完全匹配（因为我们在同一个源中查找）。在

当我尝试广播b_mat:b_mat_dist=broadcast_matrix（b_mat）时，我得到以下错误：Incompatible dimension for X and Y matrices: X.shape[1] == 56710 while Y.shape[1] == 2418

任何帮助将不胜感激！ 提前谢谢！在

这是我的代码：

import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
from pyspark.sql import SQLContext, SparkSession
from pyspark import SparkContext
from scipy.sparse import csr_matrix
vectorizer = TfidfVectorizer()

if 'sc' in locals():
    sc.stop()

sc = SparkContext("local", "Simple App")

pd.set_option('display.max_colwidth', -1)
RefB =  pd.read_excel('Ref.xlsx')
ToMatchB =  pd.read_excel('ToMatch.xlsx')

Ref = RefB['CLT_company_name']
ToMatch = ToMatchB ['Name1']

a_mat = vectorizer.fit_transform(Ref)
b_mat = vectorizer.fit_transform(ToMatch)

def find_matches_in_submatrix(sources, targets, inputs_start_index,
                              threshold=.8):
    cosimilarities = cosine_similarity(sources, targets)
    for i, cosimilarity in enumerate(cosimilarities):
        cosimilarity = cosimilarity.flatten()
        # Find the best match by using argsort()[-1]
        target_index = cosimilarity.argsort()[-1]
        source_index = inputs_start_index + i
        similarity = cosimilarity[target_index]
        if cosimilarity[target_index] >= threshold:
            yield (source_index, target_index, similarity)

def broadcast_matrix(mat):
    bcast = sc.broadcast((mat.data, mat.indices, mat.indptr))
    (data, indices, indptr) = bcast.value
    bcast_mat = csr_matrix((data, indices, indptr), shape=mat.shape)
    return bcast_mat

def parallelize_matrix(scipy_mat, rows_per_chunk=100):
    [rows, cols] = scipy_mat.shape
    i = 0
    submatrices = []
    while i < rows:
        current_chunk_size = min(rows_per_chunk, rows - i)
        submat = scipy_mat[i:i + current_chunk_size]
        submatrices.append((i, (submat.data, submat.indices, 
                                submat.indptr),
                            (current_chunk_size, cols)))
        i += current_chunk_size
    return sc.parallelize(submatrices)

a_mat_para = parallelize_matrix(a_mat, rows_per_chunk=100)
b_mat_dist = broadcast_matrix(b_mat)
results = a_mat_para.flatMap(
        lambda submatrix:
        find_matches_in_submatrix(csr_matrix(submatrix[1],
                                             shape=submatrix[2]),
                                   b_mat_dist,
                                   submatrix[0]))

Answer 1

尝试为两个TfidVectorizer对象均衡词汇表：

vect = CountVectorizer()
vocabulary =  vect.fit(Ref + ToMatch).vocabulary_
vectorizer = TfidfVectorizer(vocabulary=vocabulary)

同样基于您的目标：

^{pr2}$

对我来说是个更好的选择。在