Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Social network-based recommender systems
Schall D., Springer International Publishing, New York, NY, 2015. 126 pp. Type: Book (978-3-319227-34-4)
Date Reviewed: Jun 9 2016

This book presents a detailed summary of the author’s research related to recommender systems based on social networks. The author researched related issues during his doctoral studies at the Vienna University of Technology, which he successfully completed in 2009. The book covers a wider time span of research, resulting in more comprehensive coverage of related topics. It should be noted that this book is not some expanded version of his thesis. Instead, only some important views and results have been adopted from the dissertation for the book. The important information is that the author has a long record of successful research in the area.

Before concentrating on link formation techniques in social network-based systems, the author researched crowdsourcing and online communities, analyzed social networks, and summarized extensively the results in another book published in 2012.

This book is structured into six chapters, including a very brief final section, “Conclusion,” which only summarizes its contents.

The first section provides a brief overview of social recommender systems. The author explains his viewpoint on the wealth of recommendation techniques, structuring them into peer-based and group-based recommendations. Having identified the importance of symmetry of links, the author adds another dimension to his overview. Both peer-based and group-based recommendations take place in either undirected or directed networks. The resulting four principal cases are elaborated in chapters 2 through 5.

Chapter 2 is devoted to peer-based recommendations in directed networks. The principal research issue, as viewed by the author, is link prediction for directed graphs. Strange as it may sound, such a prediction is a basis for friendship recommendation--of course, only in the meaning of online social networks. The author presents results of his original research. His link prediction technique uses structural information as represented by oriented graphs.

Chapter 3 investigates “follow” recommendation in communities. Social collaboration platforms are typical in supporting such features as followings. A method and algorithm for who-to-follow recommendations are proposed in this chapter. Recommendation is based on a concept of user authority. Partner recommendation is the topic of chapter 4. The typical case is scientific collaboration defined as interaction that takes place within a social context. The author stresses the importance of partner reputation that is both context-dependent and time-aware. It is captured by his personalized partner activity model. Partnership is considered as a symmetric relation; links are thus undirected. The complementary case of directed links is treated in chapter 5. The author considers a scenario where there are segregated collaborative networks. Individuals who serve as intermediaries can broker information. They can also aggregate ideas that arise in different parts of a network. The importance of the concept of social trust is highlighted. Various relation metrics are discussed.

The book has several strong points. It is written by an active researcher who works in the area. His research results are highly respected by the scientific community, receiving more than a thousand citations and yielding an h-index of 20 as recorded by Google Scholar. The results form a basis for the book. It should be stressed, however, that the author presents systematically the relevant approaches to each particular topic so the reader gets a very informed overview of the current state of the art. In other words, the author’s work is consistently put into a proper context.

The book is quite brief. It contains a lot of rather technical information concentrated around particular topics. Readers may occasionally find it difficult to see a difference from a scientific journal article. On the other hand, the book lacks a few fine points to become a full-fledged monograph. In fact, almost any scientific book has some kind of introduction. Not this one. If nothing else, motivation should be clearly expressed in the beginning. The work should be put into a broader perspective. At the other end, conclusions usually give more than just a summary of the contents. The author surely has important insights to share on future directions of social network-based recommendations. But if brevity was the principal consideration, the reader will gladly overlook these few points.

I highly recommend this book to students, professionals, experts, and others interested in the potential of recommendations taking place within social networks.

Reviewer:  P. Navrat Review #: CR144493 (1608-0556)
Bookmark and Share
 
Decision Support (H.4.2 ... )
 
 
Social Networking (H.3.4 ... )
 
 
Web-Based Interaction (H.5.3 ... )
 
 
World Wide Web (WWW) (H.3.4 ... )
 
 
Group And Organization Interfaces (H.5.3 )
 
 
Systems And Software (H.3.4 )
 
Would you recommend this review?
yes
no
Other reviews under "Decision Support": Date
Mapping crime in its community setting: event geography analysis
Maltz M., Gordon A. (ed), Friedman W., Springer-Verlag New York, Inc., New York, NY, 1991. Type: Book (9780387973814)
Jul 1 1991
Database machines and decision support systems
Inmon W., QED Information Sciences, Inc., Wellesley, MA, 1991. Type: Book (9780894353352)
Nov 1 1991
Executive information systems
Burkan W., Van Nostrand Reinhold Co., New York, NY, 1991. Type: Book (9780442005672)
Oct 1 1992
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy