python酒店相似度推荐系统

import numpy as np
import pandas as pd
from nltk.corpus import stopwords
from sklearn.metrics.pairwise import linear_kernel
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.decomposition import LatentDirichletAllocation
import re
import random

import cufflinks
from plotly.offline import iplot
cufflinks.go_offline()

#加载数据集
data = pd.read_table('./Seattle_Hotels.txt',encoding="latin-1",sep = ',')
data

152 rows × 3 columns

data.shape

(152, 3)

data['desc'][100]

'On a budget in Seattle or looking for something different? The historic charm and "home away from home" atmosphere of The Baroness will be sure to make you feel like one of the family. Conveniently located on First Hill, we are proud to be part of the Virginia Mason Hospital campus and only minutes from Harborview Medical Center and Swedish Hospital. The Baroness Hotel is a great option for short or long term medical, patient or family stays. Whether you are visiting the area\'s world-class medical facilities or on a budget vacation, our goal is to ensure a wonderful stay. Guest Amenities: Complimentary Internet access, Two twin, one or two queen studios with mini fridge and microwave, Two twin or one queen suites with full kitchens, Laundry facilities available, Flat screen cable television with HBO, Complimentary local calls, Ice and vending machines located in the lobby, Coffee maker and hairdryers in all guestrooms, Room service available seven days a week from the Rhododendron Cafe, Limited wheelchair accessibility, Guest library and business center, Printing & fax services available, 100% non-smoking and pet free, Rooms are not air conditioned - fans are available, Self-parking available at Virginia Mason hospital for a fee.'

看一下酒店的主要描述信息

vec = CountVectorizer().fit(data['desc'])
bag_of_words = vec.transform(data['desc'])

bag_of_words.toarray()

array([[0, 1, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       ...,
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 1, 0, 0]], dtype=int64)

bag_of_words.shape

(152, 3200)

sum_words = bag_of_words.sum(axis =0 )

sum_words

matrix([[ 1, 11, 11, ...,  2,  6,  2]], dtype=int64)

def get_top_n_words(corpus,n=None):
    vec = CountVectorizer().fit(corpus)
    bag_of_words = vec.transform(corpus)
    sum_words = bag_of_words.sum(axis =0 )
    word_freqs = [(word,sum_words[0,idex]) for word,idex in vec.vocabulary_.items()]
    word_freq = sorted(word_freqs,key = lambda x:x[1],reverse=True)
    return word_freq[:n]

common_words = get_top_n_words(data['desc'],20)
common_words

[('the', 1258),
 ('and', 1062),
 ('of', 536),
 ('seattle', 533),
 ('to', 471),
 ('in', 449),
 ('our', 359),
 ('you', 304),
 ('hotel', 295),
 ('with', 280),
 ('is', 271),
 ('at', 231),
 ('from', 224),
 ('for', 216),
 ('your', 186),
 ('or', 161),
 ('center', 151),
 ('are', 136),
 ('downtown', 133),
 ('on', 129)]

df1 = pd.DataFrame(common_words,columns=['desc','counts'])
df1.head()

df1.groupby('desc').sum()['counts'].sort_values().iplot(kind='barh',yTitle='counts',linecolor='black',title='top 20 before remove stopwords')