Trust Networks for Recommender Systems | Date: 22 May 2011, 19:46
|
Foreword
Recommendation Systems are gaining tremendous prominence in the digital society, and are fast becoming the bastions of electronic commerce. “Will I like this book?”, “Is this a movie I can see with my kids?’, ‘Which hotel will suit me the best?”: we increasingly rely on the social aspects of the world wide web (WWW) to help us navigate through such questions in our everyday life. We are quick to judge, and even quicker to just imitate our friends and do what they did. The magic potion that casts a spell on us to imitate and even at times make irrational decisions is trust. Trust enhanced recommender systems are designed to help us to form an opinion on matters that are not entirely known to us, or evennot known at all.
The social web allows people to express their views to other users of the system. We call the resulting network a social network. There are social networks in which the users can explicitly express their opinion as trust and distrust statements. We refer to these kinds of social networks as trust networks.
The particular focus of this book, infusion of the theory of how online trust networks can be modeled and the utility of these models to enhance the quality of recommendations generated in the online recommendation systems arena is not only groundbreaking and innovative; it is likely to be the central pivot for the next generation of research in social network analysis. Think of any system where humans need subjective judgments from their
peers and seers. As you start to read the book, it will be quickly evident that issues explored in this book are the backbone of any such system. Some of these broad issues are: who you know, who you don’t know, who you trust, why you trust them, how does this trust translate, aggregate, and propagate; and how to ef?ciently and correctly identify key trust ?gures to enhance the quality of generated recommendations.
As the book notes in its introduction, the year was early 2006, the stage was set with the majority of the developed world focusing on the tremendous excitement that online communication and collaboration tools on the internet had started to foster. At the same time, the developing world was catching up very fast with the number of people getting access to communication technologies through cell-phones and low-cost internet access
points. Electronic commerce was outpacing the brick-and-mortar business in every retail sector and sites such as Epinions.com and Amazon.com were becoming the e-bazaars of the post 9/11 tumulus world where it was safer and easier to shop from home than to face the uncertainty of the physical world. The internet was enabling those that never had a voice to express their opinions freely and without the need for ?lters enforced by traditional
media. YouTube and MySpace were rede?ning individual reputations and authority in real-life more than someone’s social standing in the physical world. The key question on everyone’s mind then was: can this online world of inter-human communication be trusted?
Can trust be used to infuse con?dence in getting the most accurate, relevant and unbiased information about commodities (amazon, ebay), news (twitter), relationships (facebook), etc. Being there is only half the work. Keeping your eyes open to address these needs is the other half. The authors of this book were not only there, they also had their eyes open to see the need for addressing these issues about online trust computation to facilitate online commerce.
Considering the extraordinary ability of this book to make you think, re?ect, and to likely in?uence future research directions in enhancing our understanding of how online social networks behave, I wish you, the reader, a most wondrous journey through the folds of Trust Networks for Recommender Systems. Welcome aboard!
--Prof. Ankur M. Teredesai, Ph.D.
Institute of Technology University of Washington, Tacoma, USA
|
DISCLAIMER:
This site does not store Trust Networks for Recommender Systems on its server. We only index and link to Trust Networks for Recommender Systems provided by other sites. Please contact the content providers to delete Trust Networks for Recommender Systems if any and email us, we'll remove relevant links or contents immediately.
|
 |
|
Comments (0)
All
