IR-W2V-26.py

def ranking_ir(query):
  
  # pre-process Query
  query=query.lower()
  query=expand_contractions(query)
  query=clean_text(query)
  query=re.sub(' +',' ',query)

  # generating vector
  vector=get_embedding_w2v(query.split())

  # ranking documents
  documents=testing_corpus[['docid','title','body']].copy()
  documents['similarity']=testing_corpus['vector'].apply(lambda x: cosine_similarity(np.array(vector).reshape(1, -1),np.array(x).reshape(1, -1)).item())
  documents.sort_values(by='similarity',ascending=False,inplace=True)
  
  return documents.head(10).reset_index(drop=True)