From e26d2732745453e3add29226dd8c9a18d4b38a21 Mon Sep 17 00:00:00 2001
From: CriptDit <1487413334@qq.com>
Date: Sun, 2 Jun 2024 18:59:21 +0800
Subject: [PATCH] commit009

---
 TR/TrialRecommend/DataRequest.py | 26 ++++++++++++++-----------
 TR/TrialRecommend/Main.py        |  8 ++++----
 TR/TrialRecommend/TF_IDF.py      | 33 ++++++++++++++++++++++++++++++++
 3 files changed, 52 insertions(+), 15 deletions(-)
 create mode 100644 TR/TrialRecommend/TF_IDF.py

diff --git a/TR/TrialRecommend/DataRequest.py b/TR/TrialRecommend/DataRequest.py
index cf50f49..e645994 100644
--- a/TR/TrialRecommend/DataRequest.py
+++ b/TR/TrialRecommend/DataRequest.py
@@ -3,12 +3,16 @@ import random
 from DrissionPage import ChromiumOptions
 from DrissionPage import ChromiumPage
 import time
+import TF_IDF
+
 
 path=r"C:\Program Files\Google\Chrome\Application\chrome.exe"
 ChromiumOptions().set_browser_path(path).save()
 
 class DataRequest:
-    def __init__(self,keyword):
+    def __init__(self,keyword, user):
+        self.user=user
+        self.tf_idf='false'
         #创建文件对象
         self.success='False'
         self.resp = None
@@ -79,13 +83,16 @@ class DataRequest:
         return False
 
     def sort(self):
-        #按照喜欢人数排序
-        self.items.sort(key=lambda item:int(item['喜欢数量']),reverse=True)
-        for i in self.items:
-            if self.is_in_csv(i['文章ID']):
-                continue
-            else:
-                self.csv_writer.writerow(i)
+        if self.tf_idf=='false':
+            #按照喜欢人数排序
+            self.items.sort(key=lambda item:int(item['喜欢数量']),reverse=True)
+            for i in self.items:
+                if self.is_in_csv(i['文章ID']):
+                    continue
+                else:
+                    self.csv_writer.writerow(i)
+        else:
+            TF_IDF.tf_idf(self.items, self.user)
 
     def close(self):
         self.drive.listen.stop()
@@ -100,9 +107,6 @@ class DataRequest:
                     print('数据获取成功')
                     break
 
-                # print('[1]:',line.split(',')[1])
-                # break
-
 if __name__=='__main__':
     dataRequest=DataRequest('厦门旅游攻略')
     for i in range(1):
diff --git a/TR/TrialRecommend/Main.py b/TR/TrialRecommend/Main.py
index 07d057c..b24154f 100644
--- a/TR/TrialRecommend/Main.py
+++ b/TR/TrialRecommend/Main.py
@@ -361,11 +361,11 @@ class Mainpage:
 
     def researchPage(self):
         self.clear_csv_file()
-        rs=self.research.get()
-        if rs=='':
+        rs = self.research.get()
+        if rs == '':
             messagebox.showinfo(title='提示', message='你还没有输入任何东西！')
         else:
-            if DataRequest.DataRequest(str(self.research.get())).success=='True':
+            if DataRequest.DataRequest(str(self.research.get()), self.user).success == 'True':
 
                 self.content = ''
                 self.time = ''
@@ -383,7 +383,7 @@ class Mainpage:
                 self.max_notes_count = 0
                 self.max_iamge_count = 0
 
-                self.data_request = DataRequest.DataRequest(str(self.research.get()))
+                self.data_request = DataRequest.DataRequest(str(self.research.get()), self.user)
                 time.sleep(random.uniform(0,1))
                 self.init_notes_request()
             else:
diff --git a/TR/TrialRecommend/TF_IDF.py b/TR/TrialRecommend/TF_IDF.py
new file mode 100644
index 0000000..c2e4d97
--- /dev/null
+++ b/TR/TrialRecommend/TF_IDF.py
@@ -0,0 +1,33 @@
+import csv
+
+import pandas as pd
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
+
+def tf_idf(items, user):
+    # 数据载入
+    articles = items
+
+    df = pd.DataFrame(articles)
+
+    # 提取TF-IDF特征
+    tfidf_vectorizer = TfidfVectorizer()
+    tfidf_matrix = tfidf_vectorizer.fit_transform(df['content'])
+
+    # 计算余弦相似度
+    cosine_sim = cosine_similarity(tfidf_matrix, tfidf_matrix)
+
+    # 根据相似度推荐
+    def recommend_articles(article_id, user, cosine_sim=cosine_sim):
+        sim_scores = list(enumerate(cosine_sim[article_id - 1]))
+        sim_scores = sorted(sim_scores, key=lambda x: x[1], reverse=True)
+        sim_scores = sim_scores[1:6]  # 推荐前5个相似文章
+        article_indices = [i[0] for i in sim_scores]
+        with open('collections/{}.csv'.format(user), 'a+', encoding='utf-8',newline='') as f:
+            write=csv.DictWriter(f,fieldnames=['文章ID','content','时间','收藏数量','喜欢数量','图片资源'])
+            for j in article_indices:
+                write.writerow(items[j])
+
+
+    # 推荐与文章ID为1的旅游攻略相似的文章
+    recommend_articles(1,user)
\ No newline at end of file