from rank_bm25 import BM25Okapi

# 文本语料库
corpus = [
    "Hello there good man!",
    "It is quite windy in London",
    "How is the weather today?"
]

# 遍历每一个文档，并对文本进行切词
tokenized_corpus = [doc.split(" ") for doc in corpus]
print("tokenized_corpus-->", tokenized_corpus)

# 创建BM25对象
bm25 = BM25Okapi(tokenized_corpus)
print("bm25-->", bm25)

# 准备query
query = "windy London"
# query = "today is a nice day"
tokenized_query = query.split(" ")
print("tokenized_query-->", tokenized_query)

# 得到相似度得分
doc_scores = bm25.get_scores(tokenized_query)
print("doc_scores-->", doc_scores)