画像V3，冷启动和画像存储tag的name

eda/smart_rank/dist_update_user_portrait_service.py

@@ -40,7 +40,7 @@ def get_hot_search_words_tag():
         return []
 
 
-def get_user_service_portrait(cl_id, all_word_tags, all_tag_tag_type, all_3tag_2tag, size=10):
+def get_user_service_portrait(cl_id, all_word_tags, all_tag_tag_type, all_3tag_2tag, all_tags_name, size=10):
     try:
         db_jerry_test = pymysql.connect(host='172.16.40.158', port=4000, user='root', passwd='3SYz54LS9#^9sBvC',
                                         db='jerry_test', charset='utf8')
@@ -136,6 +136,13 @@ def get_user_service_portrait(cl_id, all_word_tags, all_tag_tag_type, all_3tag_2
             redis_client.hmset(cl_id_portrait_key2, gmkv_tag_score2_sum_dict)
             redis_client.expire(cl_id_portrait_key2, time=30 * 24 * 60 * 60)
 
+            # 标签name写redis
+            cl_id_portrait_key3 = "user:service_portrait_tags3:cl_id:" + str(cl_id)
+            gmkv_tag_score3_sum_dict = {all_tags_name[i]: gmkv_tag_score2_sum_dict[i] for i in gmkv_tag_score2_sum_dict}
+            redis_client.delete(cl_id_portrait_key3)
+            redis_client.hmset(cl_id_portrait_key3, gmkv_tag_score3_sum_dict)
+            redis_client.expire(cl_id_portrait_key3, time=30 * 24 * 60 * 60)
+
             # 写tidb，gmkv同步
             stat_date = datetime.datetime.today().strftime('%Y-%m-%d')
             replace_sql = """replace into user_service_portrait_tags (stat_date, cl_id, tag_list) values("{stat_date}","{cl_id}","{tag_list}")"""\
@@ -182,6 +189,9 @@ if __name__ == '__main__':
         # 3级tag对应的2级tag
         all_3tag_2tag = get_all_3tag_2tag()
 
+        # 标签id对应的中文名称
+        all_tags_name = get_all_tags_name()
+
         # 画像冷启动
         hot_search_words = get_hot_search_words_tag()
         hot_search_words_portrait = list()
@@ -203,8 +213,15 @@ if __name__ == '__main__':
         hot_search_words_portrait_portrait_key2 = "user:service_coldstart_tags2"
         hot_search_words_portrait_dict = {i["id"]: 0.2 for i in hot_search_words}
         gm_kv_cli.hmset(hot_search_words_portrait_portrait_key2, hot_search_words_portrait_dict)
+        redis_client.delete(hot_search_words_portrait_portrait_key2)
         redis_client.hmset(hot_search_words_portrait_portrait_key2, hot_search_words_portrait_dict)
 
+        hot_search_words_portrait_portrait_key3 = "user:service_coldstart_tags3"
+        hot_search_words_portrait3_dict = {i["keywords"]: 0.2 for i in hot_search_words}
+        redis_client.delete(hot_search_words_portrait_portrait_key3)
+        redis_client.hmset(hot_search_words_portrait_portrait_key3, hot_search_words_portrait3_dict)
+
+
         # 搜索词tag
         search_words_synonym_tags_key = "search:words:synonym:tags"
         search_words_synonym_tags_json = json.dumps(all_word_tags)
@@ -224,7 +241,7 @@ if __name__ == '__main__':
         spark.sparkContext.setLogLevel("WARN")
         spark.sparkContext.addPyFile("/srv/apps/ffm-baseline_git/eda/smart_rank/tool.py")
         device_ids_lst_rdd = spark.sparkContext.parallelize(device_ids_lst)
-        result = device_ids_lst_rdd.repartition(100).map(lambda x: get_user_service_portrait(x, all_word_tags, all_tag_tag_type, all_3tag_2tag))
+        result = device_ids_lst_rdd.repartition(100).map(lambda x: get_user_service_portrait(x, all_word_tags, all_tag_tag_type, all_3tag_2tag, all_tags_name))
         result.collect()
 
     except Exception as e:

eda/smart_rank/tool.py

@@ -174,6 +174,18 @@ def get_all_3tag_2tag():
         print(e)
 
 
+def get_all_tags_name():
+    try:
+        sql = "select id, name from api_tag where tag_type+0<'4'+0 and is_online=1"
+        mysql_results = get_data_by_mysql('172.16.30.141', 3306, 'work', 'BJQaT9VzDcuPBqkd', 'zhengxing', sql)
+        result_dict = dict()
+        for data in mysql_results:
+            result_dict[data['id']] = data['name']
+        return result_dict
+    except Exception as e:
+        print(e)
+
+
 def get_tag2_from_tag3(tag3, all_3tag_2tag, user_log_df_tag2_list):
     try:
         tag2s = []