Computational Advertising

Market and Technologies for Internet Commercial Monetization

Peng Liu, Chao Wang

• Publisher: CRC Press
• Year: 2020
• ISBN:
  9780429557729
• Edition: 2
• Topic: Computer Science
• Subtopic: Computer Science General
• Index: Computer Science

This book introduces computational advertising, and Internet monetization. It provides a macroscopic understanding of how consumer products in the Internet era push user experience and monetization to the limit. Part One of the book focuses on the basic problems and background knowledge of online advertising. Part Two targets the product, operations, and sales staff, as well as high-level decision makers of the Internet products. It explains the market structure, trading models, and the main products in computational advertising. Part Three targets systems, algorithms, and architects, and focuses on the key technical challenges of different advertising products.

Features

· Introduces computational advertising and Internet monetization

· Covers data processing, utilization, and trading

· Uses business logic as the driving force to explain online advertising products and technology advancement

· Explores the products and the technologies of computational advertising, to provide insights on the realization of personalization systems, constrained optimization, data monetization and trading, and other practical industry problems

· Includes case studies and code snippets

Preface (1)

In March 2009, I was attending CIKM’09 in Hong Kong and listened to a tutorial “Introduction to Computational Advertising” given by several scholars including Andrei Broder. I found everything fresh and new. I thought such course should be open in our university. After coming back, I got to know the situation of the young teachers around me and found no one was capable enough. Later, I learned Dr. Peng Liu is an expert in this field firstly from Weibo. I connected with him and felt happy to know that he is the one who I’ve been looking for. I invited him to give a similar lecture at Peking University, and he agreed. The lecture was scheduled one day in the summer of 2013. I designated Mr. Peng Bo, a young teacher in the laboratory, to be Liu’s assistant. One of the purposes of doing so is that I hope Peng could learn the knowledge of computational advertising while being an assistant, and one day open classes by himself in Peking University. Liu’s lectures went very well, Peng was present every time. In the fall of 2014, Peng succeeded in giving a lecture by himself. After the class, I asked him how he felt. He admitted that the content was too much and he could not grasp it well. As there was no textbook, it was difficult for both teacher and students.

In fact, when I first asked Dr. Liu to give lectures, I talked about the textbook with him. He promised to consider it, but as he has been busy with his work in the company, it would take a long time. But he did not forget it! After two years have passed, one day he sent me an email saying that the manuscript was finished and he expected me to write a preface for his book, which pleased me a lot.

It is not a big book, but it comprehensively covers the Internet-based advertising market background, product logic, and key technologies, and provides readers with a broad vision. Based on his years’ experience, the author interprets the demand for products and technologies from the perspective of market behavior, rather than focusing on technology, thus having improved the idea of this book and is therefore suitable for a wider readership, including graduate students majored in computer science. It should be said that the style of this book is different from the usual teaching materials, if directly used for teaching, there will be higher requirements for teachers, but it is an excellent teaching reference book. In particular, the main thinking line of data processing, utilization, and transaction in this book can enable computer students to learn about the concrete technical demand. The emphasis on data in the context of the Internet advertising enables readers to have a more concrete experience of the significance of big data.

There are lots of difficulties and problems in computer professional education in Chinese universities (especially for senior and graduate students). One obvious point is that the teaching content is not modern enough. This is reflected in both breadth and depth. In line with the development of industry, some important courses cannot be opened timely and effectively. Computational advertising is one of them. This situation is incompatible with the booming information technology and industry. Therefore, we welcome experts who have a thorough understanding of technology and industry and are passionate about education to participate in college teaching activities, so that our students can learn more practical skills and meet the needs of industrial development. Dr. Liu opened the course on Computational Advertising at Peking University in 2013, which sets an example in this regard. The publication of his book is also a kind of dedication in this sense.

Li Xiaoming
Professor with Department of Computer Science & Technology, Peking University

Preface (2)

All Internet companies are no strangers to the status and value of advertising monetization. At every stage of the growth of each user product, in addition to seriously solving the pain points of demand and optimizing user experience, we shall constantly evaluate the value of traffic and data, and actively discuss the strategy and product of commercial monetization. Among all kinds of commercial products, computational advertising is undoubtedly the most important.

At the early stages of product selection, development, and operation, it is very important to judge the growth space and commercial value of the product, if we can correctly evaluate the data and traffic value of the product in the future and know how to monetize these assets by utilizing advertising products. In addition, the early product promotion will use many advertising products, and the in-depth understanding of the principles of computational advertising will also be conducive to efficient marketing.

When a product is recognized by the market and has absorbed a certain number of users, it is a key step for each Internet company to actively formulate a systematic commercialization strategy and obtain cash flow in a reasonable way so to support the rapid development of product, which is a crucial stage in their growth process. If we can have a thorough understanding of the product technology of the Internet advertising market, it will be greatly helpful for the decision-making at this stage.

Although advertising technology is fairly important in the Internet industry, for a long time, there have been only a few monograph articles, and the introduction of the industry system architecture and algorithm from a global perspective lacked systematic collation and summarization. On the one hand, this is due to the rapid development of advertising market, from search bidding to programmatic trade and then to the native ad in the era of mobile Internet, and rapid product evolution has left no time for the entire industry to do a summary; on the other hand, the internal logic of advertising products is not as intuitive as that of user products. To conduct a comprehensive and thorough sorting and analysis, it requires both rich practical experience and considerable theoretical abstraction ability. Due to the lack of systematic information and incomplete talent cultivation in the Internet industry, there has been a lack of talents in advertising product technology.

Dr. Peng Liu had been working with me in Sohu for some time. Through our brief contact, I know that he has rich practical experience in monetization of media traffic and demand-side advertising products. Liu had once worked in Yahoo! Labs to do systematic research in the field of computational advertising. So Liu is the right person to do a comprehensive summary of the field. He could devote his time and energy to incorporate the product technology and business logic of computational advertising into a book that will greatly benefit the whole Internet industry.

My first impression of reading this book is that it is comprehensive and methodical: this book is an all-round introduction to computational advertising, a thorough dissection of its business logic and principles, and an in-depth discussion of its technical architecture and key algorithms. Moreover, in addition to audience targeting, CTR estimation, RTB, and other hot topics, there are detailed introductions to peripheral products and technology. I believe that readers who have carefully read this book will have an overall understanding of the whole advertising ecology, they will not only see the trees but see the forest, and readers can follow the picture and find specific ideas and even solutions in this book when they encounter various practical problems.

Another important aspect of this book, of course, is that it is the first systematic, formal publication in the field of computational advertising. I really hope that from the perspective of rational allocation of resources, the entire Internet field can gradually move toward standardization and division of labor in traffic and data monetization. This may promote Chinese Internet enterprises to get rid of vicious competition and march on the road of win-win cooperation.

In the end, I congratulate Dr. Liu for publishing this book and may it give you some enlightenment.

Wang Xiaochuan
CEO of Sogou

Preface (3)

Advertising marketing is at a turning point in its history, as the motivation and technical integration of the media has made the development of digital marketing an interesting and disturbing topic. The reason is simple: on the one hand, technology-driven digital ecology is flourishing, and programmatic marketing is getting better; on the other hand, the various concepts of advertising technology in the digital world have made marketizers confused.

It cannot be denied that the marketing industry, with its skilled professionals and data scientists helping us practice and innovate these technologies, seems to have everything it needs. Imagine sitting with one of our clients one afternoon, he might ask: What can you do for us in the future?

If technology represents the future of marketing, what is this technology? Why does technology exist? What can technology help people do? I often think that in order to break through the mists of technology, marketizers shall have a background that cuts through the blind spots of technical understanding, provides insight into the really crucial and clear attribution, and delivers the answers to these questions clearly and simply to our customers. So I’m looking for someone in the industry to explain the reality and role of advertising technology, whether it’s commercial product concepts like DSP, DMP or RTB, or technical terms like “prediction models,” “machine learning,” and “demographical targeting.”

With this expectation, I read through Dr. Liu’s book. I would like to say that I have found the answers or clues while reading this book.

Dr. Liu has rich experience in the Internet field, especially in the field of advertising monetization. From Yahoo! global research and development center to Microsoft research, and now as the chief business architect of 360, he has not only presided over the design and development of demand-side marketing products and supply-side monetization products, but capable of grasping products, systems, and algorithms, and these experiences have become the foundation of the rich content in this book.

A good advertising book doesn’t talk about trends, instead it examines the business logic in detail. A good technical book doesn’t talk about common sense, but dissects the practical issues and come up with penetrating judgments. Dr. Liu’s book is the one that crosses disciplines and has both at the same time.

I’d like to share two points of view about my reading experience. The first is the complicated digital ecology and technical rhetoric, which Dr. Liu has systematically sorted out and introduced. Even with highly professional product concepts, logic, and algorithm applications, readers from non-technical backgrounds can also establish a unified understanding of these concepts. The second is beyond the concept. The book lists classic international and domestic advertising platform products, analyzes their forms, technologies, and strategies, and depicts the interconnection and promotion between business and products. These are from the author’s years of practice and accumulation of marketing and “Internet +” thinking perspective which are more valuable. The book is full of detailed data and illustrations, reflecting Dr. Liu’s seriousness about technology and scholarship.

If you need to learn about the products and technologies of online advertising, you should do it now, open this book, and try to learn and explore.

May this book be accessible to every marketizer in digital advertising.

Li Guifen
CEO of Aegis Media Greater China

Authors

Dr. Peng Liu is senior director and chief architect of business products at Qihoo 360. He is also responsible for product and engineering for monetization of 360. After receiving his PhD from Tsinghua University in 2005, he joined Microsoft Research Asia and studied cutting-edge artificial intelligence technologies. In 2009, he participated in the founding of Yahoo! Labs Beijing as a senior scientist. He was also chief scientist of MediaV. Dr. Peng Liu is devoted to products and technologies related to big data and computational advertising. His public online course “computational advertising” has attracted more than 30,000 students on Netease.com, and has been adopted as a basic training material in many related companies. Moreover, this course has been selected by Peking University, Tsinghua University and Beihang University for their graduates.

Wang Chao received his master’s degree from Peking University, and then worked at Weibo and Autohome’s advertising department for some years. He is now a tech leader in the query recommendation group at Baidu’s portal search department. His work focuses on machine learning algorithms in computational advertising, and he has won 7^th place among 718 participants in “predict click-through rates on display ads” organized by Kaggle and Criteo. He is also interested in contributing code for open source machine learning tools such as xgboost.

1

Market and Background of Online Advertising

Chapter 1

Overview of Online Advertising

Online advertising, which is also known as network advertising or Internet advertising, refers to the ad serving via online media just as its name implies. Unlike the traditional advertising, online advertising has given birth to a technology-based and product-oriented ad serving model that targets specific user segmentation, though it was born just over a decade ago. It brings not only a new marketing channel for advertisers to accurately reach target audiences, but a means for large-scale monetization of the Internet-based free products and media. The fact is that no matter you are dealing with user products or commercial products you cannot fully understand the Internet business without a profound knowledge of online advertising. Given this, it is necessary for the Internet practitioners to spend some time figuring out the rationales and products of modern online advertising.

In addition, with regard to data application, online advertising takes the lead in large-scale and automated use of data to improve products and increase revenue. To say the least, for quite a long time in the past, online advertising was the only trade that fulfilled large-scale reve...

Table of contents

• Cover
• Endorsements
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• Figures
• Tables
• Foreword
• Preface (1)
• Preface (2)
• Preface (3)
• Authors
• PART 1: Market and Background of Online Advertising
• PART 2: Product Logic of Online Advertising
• PART 3: Key Technologies for Computational Advertising
• PART 4: Terminology and Index
• References
• INDEX