GithubHelp home page GithubHelp logo

Comments (10)

guodongxiaren avatar guodongxiaren commented on June 15, 2024

广告的定义

广告是由已确定的出资人通过各种媒介进行的有关产品(商品,服务和观点)的,通常是有偿的、有组织的、综合的、劝服性的非人员的信息传播活动

广告的主体

  • 出资人(Sponsor)即广告主(advertiser)
  • 媒介(Medium)
  • 受众(Audience)

广告的本质功能

借助某种有广泛受众的媒介的力量,完成较低成本的用户接触(reach)

广告的分类

  1. 品牌广告(Brand Awareness)
  2. 效果广告(Direct Response)

品牌广告

创造独特良好的品牌或产品形象,目的在于提升较长时期内的离线转化率

效果广告

短期内明确用户转化行为诉求的广告。用户转化行为例如:购买、注册、投票、捐款等。

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

广告的有效性模型

  1. 选择
    1. 曝光
    2. 关注
  2. 解释
    1. 理解
    2. 信息接受
  3. 态度
    1. 保持
    2. 购买

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

对广告而言,文字链的广告比图片广告的点击率要高。

但是对于推荐来说,文字链的效果远远低于图片。

广告、推荐、搜索是三个不同的领域。相同点都是 Web scale challenge。

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

在线广告的独特性

技术和计算导向

  • 数字媒体的特点是在线广告可以进行精细的受众定向
  • 技术使得广告决策和交易朝着计算驱动的方向发展

可衡量性

  • 广告的点击是效果的直接收集途径

标准化

  • 技术投放和精准定向促进了在线广告的标准化

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

在线广告的核心计算问题

image

找到用户 u,上下文(context)c 和一个合适的 广告(ad) a,三者之间的最佳匹配

从优化角度来看:

  • 特征提取:受众定向
  • 微观优化:CTR预测
  • 宏观优化:竞价市场机制
  • 受限优化:在线分配
  • 强化学习:探索与利用
  • 个性化重定向:推荐技术

从系统角度来看:

  • 候选查询:实时索引
  • 特征存储:No-SQL技术
  • 离线学习:Hadoop
  • 在线学习:流计算
  • 交易市场:实时竞价

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

在线广告的主要挑战1

大规模(Scale)

  • 百万级的页面,十亿量级的用户,需要被分析处理
  • 高并发在线投放系统
  • 时延的严格要求(ad exchange要求竞价在100ms内返回)

动态性(Dynamics)

  • 用户的关注和购物兴趣非常快速的变化

丰富的查询信息(Rich query)

  • 需要把用户和上下文中多样的信号一起用于检索广告候选

探索与发现(Explore & exploit)

  • 用户反馈数据局限于在以往投放中出现过的(a,u,c)的组合,需要主动探索未观察到的领域,以提高模型的正确性。

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

搜索、广告与推荐的比较

image

  • 广告不需要爬虫,搜索需要
  • 广告建模的数据量比搜索大
  • 搜算的索引的规模更大
  • 搜索以相关性为首要准则,广告以ROI为首要准则

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

image
点击率*点击价值
eCPM:预期获得的价值

不同的分解对应不同的市场形态

  • CPM市场:规定eCPM
  • CPC市场:动态CTR,固定click value
  • CPA/CPS/ROI市场:动态CTR与click value

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024

在线广告系统模块

image

image

from blog.

guodongxiaren avatar guodongxiaren commented on June 15, 2024
  • 离线部分比如用Hadoop做用户日志的生成,和点击预测的部分
  • 流式计算平台,做用户短时行为反馈

from blog.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.