Add Function ' recArticleList ' and Update vector.model

Analytics/analytics/def_set.py

@@ -2,6 +2,8 @@
 from Analytics.analytics.config import db
 from Analytics.analytics import wordvec
 
+from konlpy.tag import Twitter
+
 
 
 
@@ -15,23 +17,26 @@ def dbQuery_insert(sql,insertData):
         conn.close()
 
 
-def dbQuery(sql) :
+def dbQuery(sql, param) :
     result = []
     conn = db.getConnection()
     try:
         with conn.cursor() as cursor:
-            cursor.execute(sql)
+            cursor.execute(sql, param)
             _result = cursor.fetchall()
 
             for i in range(len(_result)):
                 result.append(_result[i])
 
+    except Exception as ex:  # 에러 종류
+        print('Error Break : ', ex, " in Function dbQuery ")  # ex는 발생한 에러의 이름을 받아오는 변수
+
     finally:
         conn.close()
         return result
 
 
-def article_similarity(article_1, article_2):
+def articleSimilarity(article_1, article_2):
     model = Word2Vec.load('./analytics/vector.model')
     return model.wv.similarity(article_1, article_2)
 
@@ -40,9 +45,10 @@ def article_similarity(article_1, article_2):
 
 # 기사 간의 유사도를 측정하는 function
 # 초기에 1번만 실행
-def init_btw_article_similarity():
+def initBtwArticleSimilarity():
     sql = "SELECT id,content FROM article"
-    result = dbQuery(sql)
+    param = []
+    result = dbQuery(sql, param)
     '''
     print(result[0]) # ( (id,content) )
     print(result[0][0]) # id
@@ -53,34 +59,38 @@ def init_btw_article_similarity():
         j = i + 1
         sql = "SELECT * FROM article_representation_noun WHERE id = "
         param = str(result[i][0])
-        article1_rep_noun = dbQuery(sql + param)
+        article1_rep_noun = dbQuery(sql , param)
         # print(article1_rep_noun)
         # [(20, '보드', '주행', '전동', '타이어', '제품')]
 
         while j < len(result):
             param = str(result[j][0])
-            article2_rep_noun = dbQuery(sql + param)
+            article2_rep_noun = dbQuery(sql , param)
             print("article2_rep_noun : " , article2_rep_noun)
             # [(21, '보드', '주행', '전동', '타이어', '제품')]
 
             similarity_value = 0
             loop_per_cnt = 5
 
-            for a in range(loop_per_cnt):
-                for b in range(loop_per_cnt):
-                    similarity_value += article_similarity(article1_rep_noun[0][a + 1],article2_rep_noun[0][b + 1])
+            try :
+                for a in range(loop_per_cnt):
+                    for b in range(loop_per_cnt):
+                        similarity_value += articleSimilarity(article1_rep_noun[0][a + 1],article2_rep_noun[0][b + 1])
 
-            insertQuery = "INSERT INTO article_similarity_value(article1, article2, similarity_value) VALUES ( %s , %s, %s )"
-            insertData = (str(result[i][0]) , str(result[j][0]) , str(similarity_value))
-            dbQuery_insert(insertQuery, insertData)
+                insertQuery = "INSERT INTO article_similarity_value(article1, article2, similarity_value) VALUES ( %s , %s, %s )"
+                insertData = (str(result[i][0]) , str(result[j][0]) , str(similarity_value / 25 ))
+                dbQuery_insert(insertQuery, insertData)
 
-            j += 1
+                j += 1
+            except Exception as ex:  # 에러 종류
+                print('Error Break : ', ex , " in Function initBtwArticleSimilarity")  # ex는 발생한 에러의 이름을 받아오는 변수
 
 
 
 def insertRepresentationNoun():
     selectSQL= " SELECT id, content FROM article"
-    selectResult = dbQuery(selectSQL)
+    param = []
+    selectResult = dbQuery(selectSQL, param)
 
     try:
         for i in range(len(selectResult)):
@@ -91,9 +101,88 @@ def insertRepresentationNoun():
             dbQuery_insert(sql, insertData)
 
     except Exception as ex:  # 에러 종류
-        print('Error Break : ', ex)  # ex는 발생한 에러의 이름을 받아오는 변수
+        print('Error Break : ', ex, " in Function insertRepresentationNoun ")  # ex는 발생한 에러의 이름을 받아오는 변수
 
     finally:
         print(" End :: Function insertRepresentationNoun ")
 
 
+def initTraingModel():
+    doc_ko = db.getArticle()
+    t = Twitter()
+
+    noun_token = []
+    for i in range(len(doc_ko)):
+        tokens_ko = t.morphs(doc_ko[i])
+        noun_token.append(tokens_ko)
+
+    model = Word2Vec(noun_token, min_count=1, iter=1000)
+    model.save('./analytics/vector.model')
+
+
+def recArticleList(userid) : # rec is recommend
+    selectSQL = "SELECT article_id FROM like_list WHERE user_id = %s"
+    param = (str(userid))
+    liked_list_by_user = dbQuery( selectSQL , param)
+
+    article_list = []
+
+    for i in range(len(liked_list_by_user)):
+        article_list.append(liked_list_by_user[i][0])
+
+    # 좋아요한 글들간의 유사성 값을 측정
+    size = len(article_list)
+    denominator = size * (size-1) / 2 # denominator means 분모
+
+    sum = 0
+
+    for i in range(size-1):
+        j = i + 1
+        while j < size:
+            selectSQL = "SELECT similarity_value FROM article_similarity_value WHERE article1 = %s and article2 = %s"
+            param = (article_list[i], article_list[j])
+
+            selectResult = dbQuery(selectSQL, param)
+            sum += selectResult[0][0]
+            j += 1
+
+    if sum != 0 :
+        pivot_value = sum / denominator
+    else :
+        pivot_value = 0.05
+
+
+    for i in range(size):
+        selectSQL = "SELECT article2 FROM article_similarity_value WHERE article1 = %s and similarity_value >= %s"
+        param = (article_list[i], pivot_value)
+
+        selectResult = dbQuery(selectSQL, param)
+
+        for j in range(len(selectResult)):
+            article_list.append(selectResult[j][0])
+
+    article_list = list(set(article_list))
+
+    return article_list
+
+
+
+'''
+# 새로운 Article로 모델 만들때 사용
+def_set.initTraingModel()
+'''
+
+'''
+# 글에서 대표 명사 추출
+import Analytics.analytics.def_set as a
+a.insertRepresentationNoun()
+'''
+
+'''
+# 글간의 유사도 측정을 위한 설정
+import Analytics.analytics.def_set as b
+b.initBtwArticleSimilarity()
+'''
+
+
+