# -*- coding: UTF-8 -*- # !/usr/bin/env python from kafka import KafkaConsumer import random from libs.cache import redis_client import logging from linucb.views.linucb import LinUCB import json from trans2es.models.tag import TopicTag, Tag from trans2es.models.topic import TopicHomeRecommend import traceback from django.conf import settings from libs.es import ESPerform from search.utils.common import * import libs.tools as Tools from trans2es.models.pictorial import CommunityPictorialHomeFeed from trans2es.models.portrait_stat import LikeDeviceTagStat from libs.error import logging_exception import os from search.views.tag import get_same_tagset_ids import msgpack def loads_data(data): try: result = json.loads(data) msg = True return result, msg except: result = msgpack.loads(data) msg = False return result, msg class KafkaManager(object): consumser_obj = None @classmethod def get_kafka_consumer_ins(cls, topic_name=None): if not cls.consumser_obj: topic_name = settings.KAFKA_TOPIC_NAME if not topic_name else topic_name gm_logging_name = settings.KAFKA_GM_LOGGING_TOPIC_NAME cls.consumser_obj = KafkaConsumer(bootstrap_servers=settings.KAFKA_BROKER_LIST) cls.consumser_obj.subscribe([topic_name, gm_logging_name]) return cls.consumser_obj class CollectData(object): def __init__(self): # lin tag参数 self.linucb_matrix_redis_prefix = "physical:linucb:device_id:" self.ctr_linucb_matrix_redis_prefix = "ctr_physical:linucb:device_id:" # lin推荐tag self.linucb_recommend_redis_prefix = "physical:linucb:tag_recommend:device_id:" self.ctr_linucb_recommend_redis_prefix = "ctr_physical:linucb:tag_recommend:device_id:" # 推荐帖子 self.linucb_recommend_topic_id_prefix = "physical:linucb:topic_recommend:device_id:" self.ctr_linucb_recommend_topic_id_prefix = "ctr_physical:linucb:topic_recommend:device_id:" # 推荐榜单 self.linucb_recommend_pictorial_id_prefix = "physical:linucb:pictorial_recommend:device_id:" self.ctr_linucb_recommend_pictorial_id_prefix = "ctr_physical:linucb:pictorial_recommend:device_id:" self.tag_topic_id_redis_prefix = "physical:tag_id:topic_id_list:" self.click_recommend_redis_key_prefix = "physical:click_recommend:device_id:" # 默认 self.user_feature = [0, 1] def _get_user_linucb_info(self, device_id, linucb_matrix_prefix): try: redis_key = linucb_matrix_prefix + str(device_id) # dict的key为标签ID，value为4个矩阵 redis_linucb_tag_data_dict = redis_client.hgetall(redis_key) return redis_linucb_tag_data_dict except: logging_exception() logging.error("catch exception,err_msg:%s" % traceback.format_exc()) return dict() def update_recommend_tag_list(self, device_id, user_feature=None, user_id=None, click_topic_tag_list=None, new_user_click_tag_list=[], linucb_matrix_prefix=None, linucb_recommend_tag_prefix=None, linucb_topic_ids_prefix=None, linucb_pictorial_ids_prefix=None): try: redis_linucb_tag_data_dict = self._get_user_linucb_info(device_id, linucb_matrix_prefix) if len(redis_linucb_tag_data_dict) == 0: recommend_tag_list = list(LinUCB.get_default_tag_list(user_id)) LinUCB.init_device_id_linucb_info(redis_client, linucb_matrix_prefix, device_id, recommend_tag_list) else: user_feature = user_feature if user_feature else self.user_feature linucb_tag_list = list(redis_linucb_tag_data_dict.keys()) (recommend_tag_dict, recommend_tag_set) = LinUCB.linucb_recommend_tag(device_id, redis_linucb_tag_data_dict, user_feature, linucb_tag_list) recommend_tag_list = list(recommend_tag_dict.keys()) if len(recommend_tag_list) > 0: tag_recommend_redis_key = linucb_recommend_tag_prefix + str(device_id) redis_client.set(tag_recommend_redis_key, json.dumps(recommend_tag_list)) redis_client.expire(tag_recommend_redis_key, 30 * 24 * 60 * 60) have_read_topic_id_list = Tools.get_have_read_topic_id_list(device_id, user_id, TopicPageType.HOME_RECOMMEND) have_read_lin_pictorial_id_list = Tools.get_have_read_lin_pictorial_id_list(device_id, user_id, TopicPageType.HOME_RECOMMEND) promote_recommend_topic_id_list = TopicHomeRecommend.objects.using(settings.SLAVE1_DB_NAME).filter( is_online=1).values_list("topic_id", flat=True) promote_lin_pictorial_id_list = CommunityPictorialHomeFeed.objects.using( settings.SLAVE1_DB_NAME).filter( is_deleted=0, is_online=1).values_list("pictorial_id", flat=True) have_read_topic_id_list.extend(promote_recommend_topic_id_list) have_read_lin_pictorial_id_list.extend(promote_lin_pictorial_id_list) recommend_topic_id_list = list() recommend_topic_id_list_dict = dict() recommend_topic_id_list_click = list() recommend_topic_id_list_click_dict = dict() recommend_lin_pictorial_id_list = list() if click_topic_tag_list and len(click_topic_tag_list) > 0: click_topic_tag_list_same_tagset_ids = get_same_tagset_ids(click_topic_tag_list) recommend_topic_id_list_click, recommend_topic_id_list_click_dict = ESPerform.get_tag_topic_list_dict( click_topic_tag_list_same_tagset_ids, have_read_topic_id_list, size=2) if len(recommend_topic_id_list_click) > 0: recommend_topic_id_list.extend(recommend_topic_id_list_click) recommend_topic_id_list_dict.update(recommend_topic_id_list_click_dict) # have_read_topic_id_list.extend(recommend_topic_id_list_click) # click_recommend_redis_key = self.click_recommend_redis_key_prefix + str(device_id) # click_redis_data_dict = { # "data": json.dumps(recommend_topic_id_list), # "datadict":json.dumps(recommend_topic_id_list_dict), # "cursor": 0 # } # redis_client.hmset(click_recommend_redis_key, click_redis_data_dict) tag_id_list = recommend_tag_list[0:20] pictorial_recommend_redis_key = linucb_pictorial_ids_prefix + str(device_id) topic_recommend_redis_key = linucb_topic_ids_prefix + str(device_id) # redis_topic_data_dict = redis_client.hgetall(topic_recommend_redis_key) # redis_topic_list = list() # cursor = -1 # if b"data" in redis_topic_data_dict: # redis_topic_list = json.loads(redis_topic_data_dict[b"data"]) if redis_topic_data_dict[ # b"data"] else [] # cursor = int(str(redis_topic_data_dict[b"cursor"], encoding="utf-8")) # if len(recommend_topic_id_list)==0 and cursor==0 and len(redis_topic_list)>0: # have_read_topic_id_list.extend(redis_topic_list[:2]) if len(new_user_click_tag_list) > 0: new_user_click_tag_list_same_tagset_ids = get_same_tagset_ids(new_user_click_tag_list) tag_topic_id_list, tag_topic_dict = ESPerform.get_tag_topic_list_dict( new_user_click_tag_list_same_tagset_ids, have_read_topic_id_list) recommend_lin_pictorial_id_list = ESPerform.get_tag_pictorial_id_list( new_user_click_tag_list_same_tagset_ids, have_read_lin_pictorial_id_list) else: tag_id_list_same_tagset_ids = get_same_tagset_ids(tag_id_list) tag_topic_id_list, tag_topic_dict = ESPerform.get_tag_topic_list_dict(tag_id_list_same_tagset_ids, have_read_topic_id_list) recommend_lin_pictorial_id_list = ESPerform.get_tag_pictorial_id_list(tag_id_list_same_tagset_ids, have_read_lin_pictorial_id_list) if len(recommend_topic_id_list) > 0 or len(tag_topic_id_list) > 0 or len(new_user_click_tag_list) > 0: tag_topic_id_list = recommend_topic_id_list + tag_topic_id_list tag_topic_dict.update(recommend_topic_id_list_dict) redis_data_dict = { "data": json.dumps(tag_topic_id_list), "datadict": json.dumps(tag_topic_dict), "cursor": 0 } redis_client.hmset(topic_recommend_redis_key, redis_data_dict) if len(recommend_lin_pictorial_id_list) > 0: pictorial_data_dict = { "data": json.dumps(recommend_lin_pictorial_id_list), "cursor": 0 } redis_client.hmset(pictorial_recommend_redis_key, pictorial_data_dict) return True except: logging_exception() logging.error("catch exception,err_msg:%s" % traceback.format_exc()) return False def update_user_linucb_tag_info(self, reward, device_id, tag_id, user_feature, linucb_matrix_redis_prefix): try: user_feature = user_feature if user_feature else self.user_feature return LinUCB.update_linucb_info(user_feature, reward, tag_id, device_id, linucb_matrix_redis_prefix, redis_client) except: logging_exception() logging.error("update_user_linucb_tag_info error!") return False def transfer_old_info2ctr_feature_key(self, device_id): try: # 移植老用户的lin标签参数信息到ctr特征策略 ctr_linucb_matrix_redis_prefix_key = self.ctr_linucb_matrix_redis_prefix + str(device_id) linucb_matrix_redis_prefix_key = self.linucb_matrix_redis_prefix + str(device_id) if redis_client.exists(ctr_linucb_matrix_redis_prefix_key): # 如果新策略存在lin信息，则不需要移植 return True else: if redis_client.exists(linucb_matrix_redis_prefix_key): older_device_info = redis_client.hgetall(linucb_matrix_redis_prefix_key) redis_client.hmset(ctr_linucb_matrix_redis_prefix_key, older_device_info) # 移植老用户的lin推荐标签列表信息到ctr特征策略 ctr_linucb_recommend_redis_prefix = self.ctr_linucb_recommend_redis_prefix + str(device_id) linucb_recommend_redis_prefix = self.linucb_recommend_redis_prefix + str(device_id) if not redis_client.exists(ctr_linucb_recommend_redis_prefix): if redis_client.exists(linucb_recommend_redis_prefix): older_device_info = redis_client.get(linucb_recommend_redis_prefix) redis_client.set(ctr_linucb_recommend_redis_prefix, older_device_info) # 移植老用户的lin帖子推荐队列信息到ctr特征策略 linucb_recommend_topic_id_prefix = self.linucb_recommend_topic_id_prefix + str(device_id) ctr_linucb_recommend_topic_id_prefix = self.ctr_linucb_recommend_topic_id_prefix + str(device_id) if not redis_client.exists(ctr_linucb_recommend_topic_id_prefix): if redis_client.exists(linucb_recommend_topic_id_prefix): older_device_info = redis_client.hgetall(linucb_recommend_topic_id_prefix) redis_client.hmset(ctr_linucb_recommend_topic_id_prefix, older_device_info) # 移植老用户的lin榜单推荐队列信息到ctr特征策略 linucb_recommend_pictorial_id_prefix = self.linucb_recommend_pictorial_id_prefix + str(device_id) ctr_linucb_recommend_pictorial_id_prefix = self.ctr_linucb_recommend_pictorial_id_prefix + str(device_id) if not redis_client.exists(ctr_linucb_recommend_pictorial_id_prefix): if redis_client.exists(linucb_recommend_pictorial_id_prefix): older_device_info = redis_client.hgetall(linucb_recommend_pictorial_id_prefix) redis_client.hmset(ctr_linucb_recommend_pictorial_id_prefix, older_device_info) logging.info("transfer_old_info2ctr_feature_key sucess:" + str(device_id)) return True except: logging_exception() logging.error("transfer_old_info2ctr_feature_key error!") return False def get_device_tag_ctr(self, device_id, tag_id): # 获取用户在该tag下的ctr信息 try: device_tag_ctr = LikeDeviceTagStat.objects.using(settings.SLAVE1_DB_NAME).filter( device_id=device_id, tag_id=tag_id).values("tag_ctr_30") if device_tag_ctr: device_tag_ctr_value = device_tag_ctr[0].get("tag_ctr_30", 0.001) else: device_tag_ctr_value = 0.001 logging.info("get_device_tag_ctr" + str(device_id) + str(tag_id)) return device_tag_ctr_value except: logging_exception() logging.error("get_device_tag_ctr error!") return 0.001 # 用户打标签加分 # 新增四种用户兴趣分行为 # 四种日志均为后端埋点日志 def transfer_update_recommend_tag_list(self, device_id, user_feature, user_id, tag_list, score_loop=1): if len(tag_list) > 0: is_click = 1 is_vote = 0 reward = 1 if is_click or is_vote else 0 # 移植老用户的lin信息到ctr特征策略 self.transfer_old_info2ctr_feature_key(device_id) for i in range(score_loop): for tag_id in tag_list: self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature, self.linucb_matrix_redis_prefix) # 获取tag的ctr信息 device_tag_ctr = self.get_device_tag_ctr(device_id, tag_id) user_feature_ctr = [device_tag_ctr, device_tag_ctr] self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature_ctr, self.ctr_linucb_matrix_redis_prefix) # 更新该用户的推荐tag数据，放在 更新完成user tag行为信息之后 self.update_recommend_tag_list(device_id, user_feature, user_id, click_topic_tag_list=tag_list, linucb_matrix_prefix=self.linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.linucb_recommend_pictorial_id_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, click_topic_tag_list=tag_list, linucb_matrix_prefix=self.ctr_linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.ctr_linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.ctr_linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.ctr_linucb_recommend_pictorial_id_prefix) def consume_data_from_kafka(self, topic_name=None): try: user_feature = [1, 1] kafka_consumer_obj = KafkaManager.get_kafka_consumer_ins(topic_name) while True: msg_dict = kafka_consumer_obj.poll(timeout_ms=100) for msg_key in msg_dict: consume_msg = msg_dict[msg_key] for ori_msg in consume_msg: try: raw_val_dict, msg = loads_data(ori_msg.value) if msg: logging.info(msg,ori_msg.value) else: logging.info(msg,ori_msg.value) if "type" in raw_val_dict and \ (raw_val_dict["type"] in ("on_click_feed_topic_card", "on_click_button")): click_topic_tag_list = list() device_id = "" if "on_click_feed_topic_card" == raw_val_dict["type"]: topic_id = raw_val_dict["params"]["topic_id"] device_id = raw_val_dict["device"]["device_id"] user_id = raw_val_dict["user_id"] if "user_id" in raw_val_dict else None logging.info("consume topic_id:%s,device_id:%s" % (str(topic_id), str(device_id))) # topic_tag_list = list(TopicTag.objects.using(settings.SLAVE_DB_NAME).filter(topic_id=topic_id,is_online=True).values_list("tag_id",flat=True)) # tag_query_results = Tag.objects.using(settings.SLAVE_DB_NAME).filter(id__in=topic_tag_list,is_online=True,is_deleted=False).values_list("id","collection","is_ai") # for id,collection,is_ai in tag_query_results: # if collection and is_ai: # click_topic_tag_list.append(id) topic_tag_list = list() click_results = TopicTag.objects.using(settings.SLAVE1_DB_NAME).filter( topic_id=topic_id, is_online=True).values_list("tag_id", "is_collection") for tag_id, is_collection in click_results: # topic_tag_list.append(tag_id) if is_collection: topic_tag_list.append(tag_id) tag_query_results = Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=topic_tag_list, is_online=True, is_deleted=False, is_category=False).values_list("id", "is_ai") for id, is_ai in tag_query_results: click_topic_tag_list.append(id) logging.info("positive tag_list,device_id:%s,topic_id:%s,tag_list:%s" % ( str(device_id), str(topic_id), str(click_topic_tag_list))) elif raw_val_dict["type"] == "on_click_button" and "page_name" in \ raw_val_dict["params"] and "button_name" in raw_val_dict["params"] \ and "extra_param" in raw_val_dict["params"]: if raw_val_dict["params"]["page_name"] == "search_detail" and \ raw_val_dict["params"]["button_name"] == "focus_tag": tag_name = raw_val_dict["params"]["extra_param"] device_id = raw_val_dict["device"]["device_id"] user_id = raw_val_dict["user_id"] if "user_id" in raw_val_dict else None tag_list = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter(name=tag_name, is_online=True, is_deleted=False, is_category=False).values_list( "id", flat=True)) click_topic_tag_list.extend(tag_list) logging.info( "query tag attention,positive tag_list,device_id:%s,query_name:%s,tag_list:%s" % ( str(device_id), tag_name, str(click_topic_tag_list))) logging.info("click_topic_tag_list:%s" % (str(click_topic_tag_list))) is_click = 1 is_vote = 0 reward = 1 if is_click or is_vote else 0 # 移植老用户的lin信息到ctr特征策略 self.transfer_old_info2ctr_feature_key(device_id) # 更新不同策略的lin标签参数信息 for tag_id in click_topic_tag_list: self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature, self.linucb_matrix_redis_prefix) # 获取tag的ctr信息 device_tag_ctr = self.get_device_tag_ctr(device_id, tag_id) user_feature_ctr = [device_tag_ctr, device_tag_ctr] self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature_ctr, self.ctr_linucb_matrix_redis_prefix) # 更新该用户的推荐tag数据，放在 更新完成user tag行为信息之后 if len(click_topic_tag_list) > 0: self.update_recommend_tag_list(device_id, user_feature, user_id, click_topic_tag_list=click_topic_tag_list, linucb_matrix_prefix=self.linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.linucb_recommend_pictorial_id_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, click_topic_tag_list=click_topic_tag_list, linucb_matrix_prefix=self.ctr_linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.ctr_linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.ctr_linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.ctr_linucb_recommend_pictorial_id_prefix) # elif "type" in raw_val_dict and "page_precise_exposure" == raw_val_dict["type"]: # if isinstance(raw_val_dict["params"]["exposure_cards"],str): # exposure_cards_list = json.loads(raw_val_dict["params"]["exposure_cards"]) # elif isinstance(raw_val_dict["params"]["exposure_cards"],list): # exposure_cards_list = raw_val_dict["params"]["exposure_cards"] # else: # exposure_cards_list = list() # device_id = raw_val_dict["device"]["device_id"] # user_id = raw_val_dict["user_id"] if "user_id" in raw_val_dict else None # logging.warning("type msg:%s" % raw_val_dict.get("type")) # exposure_topic_id_list = list() # for item in exposure_cards_list: # if "card_id" not in item: # continue # exposure_topic_id = item["card_id"] # logging.info( # "consume exposure topic_id:%s,device_id:%s" % (str(exposure_topic_id), str(device_id))) # if exposure_topic_id: # exposure_topic_id_list.append(exposure_topic_id) # # topic_tag_id_dict = dict() # tag_list = list() # exposure_sql_query_results = TopicTag.objects.using(settings.SLAVE_DB_NAME).\ # filter(topic_id__in=exposure_topic_id_list).\ # values_list("topic_id","tag_id","is_online","is_collection") # # if len(exposure_sql_query_results)>0: # for topic_id,tag_id,is_online,is_collection in exposure_sql_query_results: # if is_online and is_collection == 1: # tag_list.append(tag_id) # if is_online: # tag_sql_query_results = Tag.objects.using(settings.SLAVE_DB_NAME).filter( # id=tag_id).values_list("id", "collection", "is_ai") # for id, collection, is_ai in tag_sql_query_results: # if (is_ai == 1) and id not in tag_list: # tag_list.append(id) # # if topic_id not in topic_tag_id_dict: # topic_tag_id_dict[topic_id] = list() # topic_tag_id_dict[topic_id].append(tag_id) # # is_click = 0 # is_vote = 0 # # reward = 1 if is_click or is_vote else 0 # # logging.info("negative tag_list,device_id:%s,topic_tag_id_dict:%s" % ( # str(device_id), str(topic_tag_id_dict))) # for tag_id in tag_list: # self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature) # # # 更新该用户的推荐tag数据，放在 更新完成user tag行为信息之后 # self.update_recommend_tag_list(device_id, user_feature, user_id) elif "type" in raw_val_dict and "interest_choice_click_next" == raw_val_dict["type"]: if isinstance(raw_val_dict["params"]["tagid_list"], str): tagid_list = json.loads(raw_val_dict["params"]["tagid_list"]) elif isinstance(raw_val_dict["params"]["tagid_list"], list): tagid_list = raw_val_dict["params"]["tagid_list"] else: tagid_list = list() device_id = raw_val_dict["device"]["device_id"] user_id = raw_val_dict["user_id"] if "user_id" in raw_val_dict else None logging.info("interest_choice_click_next type:%s, device_id:%s, tag_ids:%s" % ( raw_val_dict.get("type", "missing type"), str(device_id), str(tagid_list))) # if len(exposure_sql_query_results)>0: if len(tagid_list) > 0: tag_query_results = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tagid_list, is_online=True, is_deleted=False, is_category=False).values_list("id", flat=True)) is_click = 1 is_vote = 0 reward = 1 if is_click or is_vote else 0 # 移植老用户的lin信息到ctr特征策略 self.transfer_old_info2ctr_feature_key(device_id) for tag_id in tag_query_results: self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature, self.linucb_matrix_redis_prefix) # 获取tag的ctr信息 device_tag_ctr = self.get_device_tag_ctr(device_id, tag_id) user_feature_ctr = [device_tag_ctr, device_tag_ctr] self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature_ctr, self.ctr_linucb_matrix_redis_prefix) # 更新该用户的推荐tag数据，放在 更新完成user tag行为信息之后 self.update_recommend_tag_list(device_id, user_feature, user_id, new_user_click_tag_list=tag_query_results, linucb_matrix_prefix=self.linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.linucb_recommend_pictorial_id_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, new_user_click_tag_list=tag_query_results, linucb_matrix_prefix=self.ctr_linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.ctr_linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.ctr_linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.ctr_linucb_recommend_pictorial_id_prefix) # 用户点击个性化push进linucb elif "type" in raw_val_dict and raw_val_dict["type"] == "on_click_push": # 后端已过滤，该tag_ids是帖子/榜单的编辑标签 if "tag_ids" in raw_val_dict["params"]: tag_ids = raw_val_dict["params"]["tag_ids"] else: # todo 客户端埋点bug，后期移除 tag_ids = raw_val_dict["params"].get("tag_ids:", []) if isinstance(tag_ids, str): tagid_list = json.loads(tag_ids) elif isinstance(tag_ids, list): tagid_list = tag_ids else: tagid_list = list() device_id = raw_val_dict["device"]["device_id"] user_id = raw_val_dict["user_id"] if "user_id" in raw_val_dict else None if len(tagid_list) > 0: tag_query_results = Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tagid_list, is_online=True, is_deleted=False, is_category=False).values_list("id", flat=True) is_click = 1 is_vote = 0 reward = 1 if is_click or is_vote else 0 # 移植老用户的lin信息到ctr特征策略 self.transfer_old_info2ctr_feature_key(device_id) for tag_id in tag_query_results: self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature, self.linucb_matrix_redis_prefix) # 获取tag的ctr信息 device_tag_ctr = self.get_device_tag_ctr(device_id, tag_id) user_feature_ctr = [device_tag_ctr, device_tag_ctr] self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature_ctr, self.ctr_linucb_matrix_redis_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, linucb_matrix_prefix=self.linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.linucb_recommend_pictorial_id_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, linucb_matrix_prefix=self.ctr_linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.ctr_linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.ctr_linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.ctr_linucb_recommend_pictorial_id_prefix) logging.info("on_click_push topic type:%s, device_id:%s, tag_ids:%s" % (raw_val_dict.get("type", "missing type"), str(device_id), str(tagid_list))) # 用户点击问题清单进linucb elif b'content' in raw_val_dict: data = json.loads(raw_val_dict[b'content']) if 'SYS' in data and 'APP' in data and 'action' in data['SYS'] and data['SYS'][ 'action'] == "venus/community/skin_check/submit_questions": device_id = data['SYS']['cl_id'] tagid_list = list(data['APP'].get('answer_tag', [])) user_id = data['SYS'].get('user_id', None) logging.info("skin_check topic type:%s, device_id:%s, answer_tag:%s" % (str(data['SYS']['action']), str(device_id), str(tagid_list))) if len(tagid_list) > 0: tag_query_results = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tagid_list, is_online=True, is_deleted=False, is_category=False).values_list("id", flat=True)) tag_query_results_multi = [i for i in tagid_list if i in tag_query_results] is_click = 1 is_vote = 0 reward = 1 if is_click or is_vote else 0 # 移植老用户的lin信息到ctr特征策略 self.transfer_old_info2ctr_feature_key(device_id) for i in range(5): for tag_id in tag_query_results_multi: self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature, self.linucb_matrix_redis_prefix) # 获取tag的ctr信息 device_tag_ctr = self.get_device_tag_ctr(device_id, tag_id) user_feature_ctr = [device_tag_ctr, device_tag_ctr] self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature_ctr, self.ctr_linucb_matrix_redis_prefix) # 更新该用户的推荐tag数据，放在 更新完成user tag行为信息之后 self.update_recommend_tag_list(device_id, user_feature, user_id, linucb_matrix_prefix=self.linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.linucb_recommend_pictorial_id_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, linucb_matrix_prefix=self.ctr_linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.ctr_linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.ctr_linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.ctr_linucb_recommend_pictorial_id_prefix) logging.info("skin_check topic type:%s, device_id:%s, tag_query_results:%s" % (str(data['SYS']['action']), str(device_id), str(tag_query_results_multi))) # 品牌问卷进linucb elif 'SYS' in data and 'APP' in data and 'action' in data['SYS'] and data['SYS'][ 'action'] == "venus/community/survey_question/submit": device_id = data['SYS']['cl_id'] tagid_list = list(data['APP'].get('answer_tag', [])) user_id = data['SYS'].get('user_id', None) logging.info("survey_question type:%s, device_id:%s, answer_tag:%s" % (str(data['SYS']['action']), str(device_id), str(tagid_list))) if len(tagid_list) > 0: tag_query_results = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tagid_list, is_online=True, is_deleted=False, is_category=False).values_list("id", flat=True)) tag_query_results_multi = [i for i in tagid_list if i in tag_query_results] is_click = 1 is_vote = 0 reward = 1 if is_click or is_vote else 0 # 移植老用户的lin信息到ctr特征策略 self.transfer_old_info2ctr_feature_key(device_id) for i in range(5): for tag_id in tag_query_results_multi: self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature, self.linucb_matrix_redis_prefix) # 获取tag的ctr信息 device_tag_ctr = self.get_device_tag_ctr(device_id, tag_id) user_feature_ctr = [device_tag_ctr, device_tag_ctr] self.update_user_linucb_tag_info(reward, device_id, tag_id, user_feature_ctr, self.ctr_linucb_matrix_redis_prefix) # 更新该用户的推荐tag数据，放在 更新完成user tag行为信息之后 self.update_recommend_tag_list(device_id, user_feature, user_id, linucb_matrix_prefix=self.linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.linucb_recommend_pictorial_id_prefix) self.update_recommend_tag_list(device_id, user_feature, user_id, linucb_matrix_prefix=self.ctr_linucb_matrix_redis_prefix, linucb_recommend_tag_prefix=self.ctr_linucb_recommend_redis_prefix, linucb_topic_ids_prefix=self.ctr_linucb_recommend_topic_id_prefix, linucb_pictorial_ids_prefix=self.ctr_linucb_recommend_pictorial_id_prefix) logging.info("survey_question type:%s, device_id:%s, tagid_list:%s" % (str(data['SYS']['action']), str(device_id), str(tag_query_results_multi))) # 首页搜索精准匹配标签关键字进linucb elif 'SYS' in data and 'APP' in data and 'action' in data['SYS'] and \ "api/v1/cards/topic" in str(data['SYS'].get('action', "")): logging.info("action=api/v1/cards/topic") tag_name = data["APP"].get("query", []) tag_list = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( name=tag_name).values_list("id")) device_id = data["SYS"]["cl_id"] user_id = data['SYS'].get('user_id', None) self.transfer_update_recommend_tag_list(device_id, user_feature, user_id, tag_list, 5) logging.info( "api/v1/cards/topic,device_id:%s,tag_list:%s" % (str(device_id), str(tag_list))) # (客户端创建回答,后台创建回答或修改回答关联标签关键字进linucb elif 'SYS' in data and 'APP' in data and 'action' in data['SYS'] and \ ("venus/community/topic/create" in str(data['SYS'].get('action',"")) or "venus/sun/topic/edit" in str(data['SYS'].get('action',"")) ): action = str(data['SYS'].get('action', '')) logging.info("action=%s" % (action)) tag_ids = list(data["APP"].get("tag_ids", [])) tag_list = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tag_ids, is_online=True, is_deleted=False, is_category=False).values_list("id")) device_id = data["SYS"]["cl_id"] user_id = data['SYS'].get('user_id', None) self.transfer_update_recommend_tag_list(device_id, user_feature, user_id, tag_list, 10) logging.info("%s,device_id:%s,tag_list:%s" % (action, str(device_id), str(tag_list))) # 创建问题关注标签关键字进linucb elif 'SYS' in data and 'APP' in data and 'action' in data['SYS'] and \ "venus/sun/pictorial/edit" in str(data['SYS'].get('action',"")): action = str(data['SYS'].get('action', '')) logging.info("action=%s" % (action)) tag_ids = list(data["APP"].get("tag_ids", [])) tag_list = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tag_ids, is_online=True, is_deleted=False, is_category=False).values_list("id")) device_id = data["SYS"]["cl_id"] user_id = data['SYS'].get('user_id', None) self.transfer_update_recommend_tag_list(device_id, user_feature, user_id, tag_list, 20) logging.info("%s,device_id:%s,tag_list:%s" % (action, str(device_id), str(tag_list))) # kyc最后一题 elif 'SYS' in data and 'APP' in data and 'action' in data['SYS'] and \ "venus/community/survey_question/record_kyc_last_question" in str(data['SYS'].get('action',"")): action = str(data['SYS'].get('action', '')) tag_ids = list(data["APP"].get("tag_ids", [])) logging.info('action:%s,tag_list:%s' % (action,str(tag_ids))) tag_query_results = list(Tag.objects.using(settings.SLAVE1_DB_NAME).filter( id__in=tag_ids, is_online=True, is_deleted=False, is_category=False).values_list("id")) tag_query_results = [i[0] for i in tag_query_results] logging.info('action:%s,mysql query taglist:%s'% (action,str(tag_query_results))) tag_query_results_multi = [i for i in tag_ids if i in tag_query_results] device_id = data["SYS"]["cl_id"] user_id = data['SYS'].get('user_id', None) self.transfer_update_recommend_tag_list(device_id, user_feature, user_id, tag_query_results_multi, 5) logging.info("action:%s,device_id:%s,tag_list:%s" % (action, str(device_id), str(tag_query_results_multi))) else: if msg: logging.warning( "unknown type msg:%s" % raw_val_dict.get("type", "missing type")) except: logging_exception() logging.error("catch exception,err_msg:%s" % traceback.format_exc()) # 假设数据库连接异常，强制退出程序，supervisor重启linub os._exit(0) except: logging_exception() logging.error("catch exception,err_msg:%s" % traceback.format_exc()) return False