FRANCESCO RICCI LIOR ROKACH BRACHA SHAPIRA PAUL B. KANTOR EDITORS RECOMMENDER SYSTEMS HANDBOOK Springer
Recommender Systems handbook
Recommender Systems Handbook
francesco ricci· Lior rokach· Bracha Shapira Paul b. Kantor Editors Recommender systems Handbook ②S ringer
Francesco Ricci · Lior Rokach · Bracha Shapira · Paul B. Kantor Editors Recommender Systems Handbook 123
editors Free University of Bozen-Bolzano aculty of Computer Science laz/ Dept. Information Systems 39100 bolar ering 84105 Beer-Sheva fricciaunibz. it work@bguacil Paul B. Kantor union University of the School of Communication, Dept. Information Systems Information Library studies Huntington Street 4 eer -Sheva 08901-1071 New Brunswick srael New Jersey apira SCILS Bldg kantor @scils. rutgers. edu isbn 978- New York I 35819-7 e-lSBN978-0-387-85820-3 DOI10.10 0-387-85820-3 Dordrecht Heidelberg London Library of Congress Control Number: 2010937590 onnection with any form of information storage and retrieval, electronic adaptation ftware, or by similar or dissimilar methodology now known or hereafter developed is forbidde The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion hether or not they are subject to proprietary rights
Lior Rokach Ben-Gurion University of the Negev Dept. Information Systems Engineering 84105 Beer-Sheva Israel liorrk@bgu.ac.il Paul B. Kantor Rutgers University School of Communication, Information & Library Studies Huntington Street 4 08901-1071 New Brunswick New Jersey SCILS Bldg. USA ISBN 978-0-387-85819-7 e-ISBN 978-0-387-85820-3 DOI 10.1007/978-0-387-85820-3 Springer New York Dordrecht Heidelberg London c Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com) Library of Congress Control Number: 2010937590 Editors Francesco Ricci Free University of Bozen-Bolzano Faculty of Computer Science Piazza Domenicani 3 39100 Bolzano Italy fricci@unibz.it Bracha Shapira Ben-Gurion University of the Negev Dept. Information Systems Engineering Beer-Sheva Israel bshapira@bgu.ac.il kantor@scils.rutgers.edu
Dedicated to our families in appreciation for their patience and support during the preparation of this handbook ER LR B S PK
Dedicated to our families in appreciation for their patience and support during the preparation of this handbook. F.R. L.R. B.S. P.K
Preface Recommender Systems are software tools and techniques providing suggestions for items to be of use to a user. The suggestions provided are aimed at supporting their users in various decision-making processes, such as what items to buy, what music to listen, or what news to read. Recommender systems have proven to be valu able means for online users to cope with the information overload and have become one of the most powerful and popular tools in electronic commerce Correspondingly, various techniques for recommendation generation have been proposed and during the last decade, many of them have also been successfully deployed in commercial environments Development of recommender systems is a multi-disciplinary effort which in- volves experts from various fields such as Artificial intelligence, Human Computer Interaction, Information Technology, Data Mining, Statistics, Adaptive User Inter aces, Decision Support Systems, Marketing, or Consumer Behavior. Recommender Systems Handbook: A Complete Guide for Research Scientists and Practitioners aims to impose a degree of order upon this diversity by presenting a coherent and unified repository of recommender systems' major concepts, theories, methodolo- gies, trends, challenges and applications. This is the first comprehensive book which is dedicated entirely to the field of recommender systems and covers several aspects of the major techniques. Its informative, factual pages will provide researchers, stu- dents and practitioners in industry with a comprehensive, yet concise and con- venient reference source to recommender systems. The book describes in detail the classical methods, as well as extensions and novel approaches that were recently in- troduced. The book consists of five parts: techniques, applications and evaluation of recommender systems, interacting with recommender systems, recommender sys- tems and communities, and advanced algorithms. The first part presents the most popular and fundamental techniques used nowadays for building recommender sys- tems, such as collaborative filtering, content-based filtering, data mining methods and context-aware methods. The second part starts by surveying techniques and ap proaches that have been used to evaluate the quality of the recommendations. Then deals with the practical aspects of designing recommender systems, it describes de- sign and implementation consideration, setting guidelines for the selection of the
Preface Recommender Systems are software tools and techniques providing suggestions for items to be of use to a user. The suggestions provided are aimed at supporting their users in various decision-making processes, such as what items to buy, what music Development of recommender systems is a multi-disciplinary effort which involves experts from various fields such as Artificial intelligence, Human Computer Interaction, Information Technology, Data Mining, Statistics, Adaptive User Interfaces, Decision Support Systems, Marketing, or Consumer Behavior. Recommender Systems Handbook: A Complete Guide for Research Scientists and Practitioners aims to impose a degree of order upon this diversity by presenting a coherent and unified repository of recommender systems’ major concepts, theories, methodologies, trends, challenges and applications. This is the first comprehensive book which is dedicated entirely to the field of recommender systems and covers several aspects of the major techniques. Its informative, factual pages will provide researchers, stuclassical methods, as well as extensions and novel approaches that were recently introduced. The book consists of five parts: techniques, applications and evaluation of recommender systems, interacting with recommender systems, recommender systems and communities, and advanced algorithms. The first part presents the most popular and fundamental techniques used nowadays for building recommender systems, such as collaborative filtering, content-based filtering, data mining methods and context-aware methods. The second part starts by surveying techniques and approaches that have been used to evaluate the quality of the recommendations. Then deals with the practical aspects of designing recommender systems, it describes design and implementation consideration, setting guidelines for the selection of the vii to listen, or what news to read. Recommender systems have proven to be valuable means for online users to cope with the information overload and have Correspondingly, various techniques for recommendation generation have been proposed and during the last decade, many of them have also been successfully deployed in commercial environments. become one of the most powerful and popular tools in electronic commerce. dents and practitioners in industry with a comprehensive, yet concise and convenient reference source to recommender systems. The book describes in detail the
Preface nore suitable algorithms. The section continues considering aspects that may affect the design and finally, it discusses methods, challenges and measures to be applied for the evaluation of the developed systems. The third part includes papers dealing with a number of issues related to the presentation, browsing, explanation and vi- sualization of the recommendations, and techniques that make the recommendation process more structured and conversational The fourth part is fully dedicated to a rather new topic, which is however rooted in the core idea of a collaborative recommender, i.e., exploiting user generated content of various types to build new types and more credible recommendations Finally the last section collects a few papers on some advanced topics, such as the exploitation of active learning principles to guide the acquisition of new knowl- edge, techniques suitable for making a recommender system robust against attacks of malicious users, and recommender systems that aggregate multiple types of user Backs and preferences to build more reliable recommendations We would like to thank all authors for their valuable contributions We would like to express gratitude for all reviewers that generously gave comments on drafts or counsel otherwise. We would like to express our special thanks to Susan Lagerstrom- Fife and staff members of Springer for their kind cooperation throughout the pro- duction of this book. Finally, we wish this handbook will contribute to the growth of this subject, we wish to the novices a fruitful learning path, and to those more ex- perts a compelling application of the ideas discussed in this handbook and a fruitful development of this challenging research area. francesco ricci Lior rokach Bracha Shapira Paul B. Kantor
viii Preface more suitable algorithms. The section continues considering aspects that may affect the design and finally, it discusses methods, challenges and measures to be applied for the evaluation of the developed systems. The third part includes papers dealing with a number of issues related to the presentation, browsing, explanation and visualization of the recommendations, and techniques that make the recommendation process more structured and conversational. The fourth part is fully dedicated to a rather new topic, which is however rooted in the core idea of a collaborative recommender, i.e., exploiting user generated content Finally the last section collects a few papers on some advanced topics, such as the exploitation of active learning principles to guide the acquisition of new knowledge, techniques suitable for making a recommender system robust against attacks of malicious users, and recommender systems that aggregate multiple types of user feedbacks and preferences to build more reliable recommendations. We would like to thank all authors for their valuable contributions. We would like to express gratitude for all reviewers that generously gave comments on drafts or counsel otherwise. We would like to express our special thanks to Susan LagerstromFife and staff members of Springer for their kind cooperation throughout the production of this book. Finally, we wish this handbook will contribute to the growth of this subject, we wish to the novices a fruitful learning path, and to those more experts a compelling application of the ideas discussed in this handbook and a fruitful Francesco Ricci Lior Rokach Bracha Shapira May 2010 Paul B. Kantor of various types to build new types and more credible recommendations. development of this challenging research area
Contents 1 Introduction to Recommender Systems Handbook francesco Ricci, Lior Rokach and Bracha Shapira 1.1 Introduction 1. 2 Recommender Systems Function 4 1.3 Data and Knowledge sources 1.4 Recommendation Techniques 1.5 Application and Evaluation 1.6 Recommender Systems and Human Computer Interaction 1.6.1 Trust, Explanations and Persuasiveness 0478 1.6.2 Conversational Systems 1.6.3 Visualization 1.7 Recommender Systems as a Multi-Disciplinary Field Emerg 1.8.1 Emerging Topics Discussed in the Handbook 18. 2 Challenges References Part I Basic Techniques 2 Data Mining Methods for Recommender Systems Xavier Amatriain, Alejandro Jaimes, Nuria Oliver, and Josep M. Puj 2.1 Introduction 2.2 Data Preprocessing 2.2.1 Similarity Measures 2.2.3 Reducing Dimensionality 2.2.4 Denoiser 3 Classification 2.3. 1 Nearest Neighbors ecision irees 2.3.3 Ruled-based Classifiers
Contents 1 Introduction to Recommender Systems Handbook . . . . . . . . . . . . . . . . 1 Francesco Ricci, Lior Rokach and Bracha Shapira 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Recommender Systems Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Data and Knowledge Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.4 Recommendation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.5 Application and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.6 Recommender Systems and Human Computer Interaction . . . . . . . 17 1.6.1 Trust, Explanations and Persuasiveness . . . . . . . . . . . . . . . 18 1.6.2 Conversational Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.6.3 Visualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1.7 Recommender Systems as a Multi-Disciplinary Field . . . . . . . . . . . 21 1.8 Emerging Topics and Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 1.8.1 Emerging Topics Discussed in the Handbook . . . . . . . . . . 23 1.8.2 Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Part I Basic Techniques 2 Data Mining Methods for Recommender Systems . . . . . . . . . . . . . . . . 39 Xavier Amatriain, Alejandro Jaimes, Nuria Oliver, and Josep M. Pujol 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.2 Data Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 2.2.1 Similarity Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 2.2.2 Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 2.2.3 Reducing Dimensionality . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.2.4 Denoising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 2.3 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2.3.1 Nearest Neighbors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2.3.2 Decision Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 2.3.3 Ruled-based Classifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 ix
