Recommendation systems: Technology, evaluation, and efficient algorithms

This article is a computer Quality Pre-sale recommendation >>>>Recommendation System: Technology, evaluation and efficient algorithms

Content Introduction

This book brings together the theoretical results and practical experience of experts and scholars in different fields, comprehensively introduces the main concepts, theories, trends, challenges and applications of recommender systems, and explains in detail how to support user decision making, planning and purchasing process. The book contains both the classical method and some new research results, covering artificial intelligence, human-computer interaction, information technology, data mining, statistics, adaptive user interface, decision support system, market and customer behavior, whether engaged in technology development, or engaged in product marketing readers, can benefit from it.
This book can be divided into five parts, a total of 25 chapters. The 1th chapter is an overview, the system introduces the concept of recommendation system, function, application field and the current application process encountered problems and challenges. The first part (the 2nd to 7th chapter) introduces the most commonly used technologies for building recommender systems, such as collaborative filtering, content-based data mining methods, and context-sensitive methods. The second part (chapter 8th to 12th) outlines the techniques and methodologies that have been used to evaluate the quality of recommendations, covering recommendations for system design and practice, describing the considerations for designing and implementing recommender systems, providing guidelines for selecting more appropriate algorithms, and evaluating the methodology, challenges, and evaluation indicators used to develop recommender systems. The third part (chapter 13th to 17th) discusses several issues such as how recommender systems are presented, browsed, interpreted, and visualized, and this section discusses techniques that make the recommendation process more structured and interoperable. Part IV (18th to 21st) discusses the use of various types of user generated content (UGC, such as tags, search queries, trust evaluation, etc.) to produce a novel type and more credible recommendation results. Part V (22nd to 25th) discusses advanced topics in recommender systems, such as exploring the principles of active learning to guide access to new knowledge, the appropriate technology to prevent referral systems from being attacked by malicious users, and how to integrate multiple types of user feedback and user preferences information to construct more reliable recommender systems.  

Translator sequence

Hu Cong (HU Head): This book is really a thick book, at the beginning of the translation of the time, repeatedly browsing the contents of the 25 chapters of the content, an upsurge of excitement, excited, hope to put down all the hands to study it. In the big Data wave, the recommendation system as a major branch, gradually by the domestic counterparts, but has been suffering from the lack of information available. Presumably many of the students who have just entered this line are asking "is there a full set of recommended system learning resources", "What are the recommended system algorithms in the industry", and so on, this book can bring a good answer. This book from the overview to the details, from the algorithm theory to the industry application, the layer of analysis of the recommendation system of technical details and application direction, the book knowledge covers a broader but without losing details, I believe it will become a recommendation system enthusiasts and practitioners in the hands of a reference book.
Wu Bing: The Netflix competition has powerfully pushed forward the extensive and in-depth research on the recommender system, which has led to new research hotspots. However, there is still a lack of more complete Chinese popular science books in China at present. Therefore, the emergence of this book can effectively fill this gap, and further promote the research and application of the recommendation system. In general, this book can be used as an introductory book for professional researchers in recommendation systems, as well as engineering staff for general application recommendation System technology. The love and research of the recommender system is one of the important reasons why I have the privilege to participate in the translation of this book, and therefore I have met many like-minded small partners. I am very grateful to them for their help and guidance, and their discussions and exchanges have benefited me a lot. At the same time, also want to thank Yan and Shanfan teachers, they give up valuable rest time to review and proofread this book, rigorous work style and attitude admirable.
Ding Bin (because): in the post-graduate period, the recommendation system will become an important method to solve the data overload after the search. The opportunity to participate in the translation of the book is very honored, although during the internship, job hunting, graduation and other series of events, but in the support of friends and help, or insist on the completion of their own responsible part of the translation work. Through the translation of this book, not only to enhance their understanding of the recommendation system, but also enhance the recommendation system in the Big data era of the role of confidence! Special thanks to Li Yanmin for the help of the translation of the book, as well as Lanci teacher carefully responsible for the audit and proofreading work! Due to the level of translation and time constraints, the translation of this book inevitably exist shortcomings, welcome readers to criticize the friends, thank you!
Wang Sheli: This is a "special" and "Great" book, "Special" in that it aggregates all the knowledge of the areas covered by the Recommender system and displays it methodically as the first book in the field of recommendation systems; "Great" is that it condenses the wisdom of many people, For example, the long reference catalogue behind each chapter reflects this. This book is a good starting point for beginners, not only can systematically understand the components of the recommendation system and how to proceed to design a recommendation system, so that beginners will not find it impossible, and the book also describes the challenge of the recommendation system, may be in the process of reading a sudden inspiration, find their own research direction. In addition, very fortunate to be able to participate in the translation of the book, very grateful to complete the translation of the book's friends, but also hope that readers of our work to provide valuable advice, thank you!
Li Yanmin: The content and importance of this book is self-evident. During the 15-month translation process, a total of 49 students participated in the translation and audit, during the promotion to the father, graduated from work, married into a happy paradise, but also began to start a business, here sincerely thank them and their families can participate in and support our translation. At the same time special thanks to Lanci teacher can help me to review together. I hope this book will help readers. Readers have questions or suggestions that can be discussed in this book forum (www.rec-sys.net).
Lanci: Believe that this book should be the aspiration of every recommendation system enthusiast, its profound and extensive, but its voluminous is also daunting. Fortunately, and do not work hard, just learn the small partners to complete this wish, over the mountains, income is the view of the Yimapingchuan, the ladder, can help more small partners siege of the village, drawbacks. The book is finally available, and we awaken it with focus and tenacity because we have a hunch that its magic will summon more engineers who are interested in it, and that the power of technology will change the world.

Objective

Referral systems are software tools and techniques for recommending the required items to the user. The recommendations are designed to support users through a variety of decision-making processes, such as what to buy, what songs to listen to, or what news to read. Referral System is a valuable method for online users to handle information overload, and becomes the most powerful and popular tool in the field of e-commerce. As a result, a wide range of recommended technologies have been proposed and many of these methods have been successfully applied in the business world over the past 10 years.
the development of recommendation systems requires multidisciplinary support, involving expert knowledge from various fields such as artificial intelligence, human-computer interaction, information retrieval, data mining, data statistics, adaptive user interfaces, decision support systems, marketing or consumer behavior. Based on this diversity, the book aims to help readers sort through the differences by demonstrating a coherent and unified knowledge system of the main concepts, theories, methodologies, trends, challenges, and applications of the recommender system. This is the first book to fully elaborate on the recommendation system, covering many aspects of the main technology. The wealth of information and practical content in this book provides a comprehensive but concise reference source for researchers, students, and practitioners in the industry on referral systems. This book not only introduces the classical method in detail, but also introduces the new method and its extension in the recent introduction. The book consists of five parts: technology, application and evaluation of recommender systems, interaction of referral systems, referral systems and community and advanced algorithms. The first section shows the most popular and basic technologies for building recommender systems today, such as collaborative filtering, content-based filtering, data mining methods, and context-aware approaches. The second part first introduces the research techniques and methods used to evaluate the quality of the recommendation, and secondly explains the practical aspects of the design recommendation system, such as the design and implementation considerations, the selection of a more suitable algorithm for the environment Guide; the related aspects that may affect the design are discussed again, and the methods, challenges and measures of the application in the evaluation of the developed system The third part includes articles that explore a range of issues, including recommended presentation, browsing, interpretation, and visualization, and techniques for making the referral process more structured and convenient.
the fourth part is completely focused on a whole new topic, but the topic is based on the main idea of filtering recommendations, such as using various types of content generated by users to build a recommender system with new types and more credibility.
Part fifth collects articles on high-level topics, such as the use of active learning techniques to guide new knowledge learning, the right technology to build robust recommender systems that can withstand malicious user attacks, and the generation of more credible recommender systems with multiple user feedback and preferences.
we would like to thank all the authors who have contributed to this book. Thanks to all the reviewers for their generous comments and suggestions. Thanks to the members of Susan Lagerstromfife and Springer for their cooperation in the process of writing this book. Finally, we hope that this handbook will contribute to the development of this discipline, provide a fruitful learning program for novices, and inspire more professionals interested in participating in the topics discussed in this book, so that this challenging field can be fruitful and make great strides.
Francesco Ricci
Lior Rokach
Bracha Shapira
Paul B.kantor
May 2010

Copyright NOTICE: This article for Bo Master original article, without Bo Master permission not reproduced.

Recommendation systems: Technology, evaluation, and efficient algorithms