Detection of Obfuscated Attacks in Collaborative Recommender Systems

This paper discusses the various types of obfuscated attacks and their characteristics. The vulnerability of collaborative recommender systemshas been well established; particularly to reverse-engineered attacks designed to bias the system in an attacker’s favor. Recent research has begun to examine detection schemes to recognize and defeat the effects of known attack models. This paper proposes some ways of detecting an avoiding these attacks. We explore empirically the impact of these obfuscated attacks against systems with and without detection, and discuss alternate approaches to reducingthe effectiveness of such attacks.

link: http://maya.cs.depaul.edu/~mobasher/papers/wmbsb-ecai-ws06.pdf

Lies and Propaganda: Detecting Spam Users in Collaborative Filtering

This paper talks about the different ways of detecting spam users. It discusses and explains various algorithms for the same. Lies and Propaganda may be spread bya malicious user who may have an interest in promoting,or downplaying the popularity of an item. By doing thissystematically, with eithermultiple identities, or by involving more people, a few malicious user votes and profiles can be injected into a collaborative recommender system. In this work, provide a simple unsupervised algorithm is provided, which exploitsstatistical properties of effective spam profiles to provide a highly accurate and fast algorithm for detecting spam.

link to paper: http://delivery.acm.org/10.1145/1220000/1216307/p14-mehta.pdf?key1=1216307&key2=7374527021&coll=GUIDE&dl=GUIDE&CFID=62245235&CFTOKEN=96995055

A Graph-based Recommender System for Digital Library

While most existing
recommender systems rely either on a content-based
approach or a collaborative approach to make
recommendations, there is potential to improve
recommendation quality by using a combination of both
approaches (a hybrid approach). In this paper, we report how
we tested the idea of using a graph-based recommender
system that naturally combines the content-based and
collaborative approaches. Due to the similarity between our
problem and a concept retrieval task, a Hopfield net
algorithm was used to exploit high-degree book-book, useruser
and book-user associations. Sample hold-out testing and
preliminary subject testing were conducted to evaluate the
system, by which it was found that the system gained
improvement with respect to both precision and recall by
combining content-based and collaborative approaches.
However, no significant improvement was observed by
exploiting high-degree associations.

link to paper: http://dlist.sir.arizona.edu/428/01/huang4.pdf

Foafing the Music: Bridging the semantic gap in music recommendation

In this paper there is an overview of the Foafing the Music system. The system uses the Friend of a Friend (FOAF) and RDF Site Summary (RSS) vocabularies for recommending music to a user, depending on the user's musical tastes and listening habits. Music information(new album releases, podcast sessions, audio from MP3 blogs, related artists' news and upcoming gigs) is gathered from thousands ofRSS feeds.

The presented system provides music discovery by means of: user profiling (defined in the user's FOAF description), context based information (extracted from music related RSS feeds) and content based descriptions(extracted from the audio itself), based on a common ontology (OWLDL) that describes the music domain.The system is available at: http://foafing-the-music.iua.upf.edu


The link to the paper :http://iswc2006.semanticweb.org/items/swchallenge_celma.pdf

Application of Dimensionality Reduction in Recommender System

This paper presents two different experiments where
one technology called Singular Value Decomposition (SVD)
is explored to reduce the
dimensionality of recommender system databases.
Each experiment compares the quality of a
recommender system using SVD with the quality of a
recommender system using collaborative filtering.
The first experiment compares the effectiveness of
the two recommender systems at predicting consumer
preferences based on a database of explicit ratings of
products. The second experiment compares the
effectiveness of the two recommender systems at
producing Top-N lists based on a real-life customer
purchase database from an E-Commerce site.

Please see the link:
http://www.grouplens.org/papers/pdf/webKDD00.pdf

Hybrid Recommender Systems

Recommender systems represent user preferences for the purpose of suggesting
items to purchase or examine. They have become fundamental applications in
electronic commerce and information access, providing suggestions that
effectively prune large information spaces so that users are directed toward those
items that best meet their needs and preferences. A variety of techniques have
been proposed for performing recommendation, including content-based,
collaborative, knowledge-based and other techniques. To improve performance,
these methods have sometimes been combined in hybrid recommenders. This
paper surveys the landscape of actual and possible hybrid recommenders, and
introduces a novel hybrid, EntreeC, a system that combines knowledge-based
recommendation and collaborative filtering to recommend restaurants. Further,
we show that semantic ratings obtained from the knowledge-based part of the
system enhance the effectiveness of collaborative filtering.


All of the known recommendation techniques have strengths and weaknesses, and manyresearchers have chosen to combine techniques in different ways. This article surveys the different recommendationtechniques being researched — analyzing them in terms of the data that supports the recommendations and thealgorithms that operate on that data — and examines the range of hybridization techniques that have been proposed.This analysis points to a number of possible hybrids that have yet to be explored. Finally, how adding ahybrid with collaborative filtering improved the performance of our knowledge-based recommender system Entree is discussed.

This links to the paper:

http://www-mmt.inf.tu-dresden.de/Lehre/Archiv/Sommersemester_04/Hauptseminar/papers/burke2002.pdf

FilmTrust: Movie Recommendations using Trust

In this paper, we present FilmTrust, a website that
integrates Semantic Web-based social networks,
augmented with trust, to create predictive movie
recommendations. We show how these
recommendations are more accurate than other
techniques in certain cases, and discuss this
technique as a mechanism of Semantic Web
interaction.

link: http://trust.mindswap.org/papers/CCNC2006.pdf

Attacking recommender systems

A good (but very long) paper by Mobasher et al., "Towards trustworthy recommender systems: An analysis of attack models and algorithm robustness" (PDF), explores a variety of ways of spamming or otherwise manipulating recommendation systems.Some excerpts:


An attack against a collaborative filtering recommender system consists of a set of attack profiles, each contained biased rating data associated with a fictitious user identity, and including a target item, the item that the attacker wishes the system to recommend more highly (a push attack), or wishes to prevent the system from recommending (a nuke attack).Previous work had suggested that item-based collaborative filtering might provide significant robustness compared to the user-based algorithm, but, as this paper shows, the item-based algorithm also is still vulnerable in the face of some of the attacks we introduced.


The paper lists several types of attacks, suggests several ways to detect attacks, tests several attacks using the GroupLens movie data set, and concludes that "item-based proved far more robust overall" but that "a knowledge-based / collaborative hybrid recommendation algorithm .... [that] extends item-based similarity by combining it with content based similarity .... [seems] likely to provide defensive advantages for recommender systems."It is worth noting that spam is a much worse problem with winner-take-all systems that show the most popular or most highly rated articles (like Digg). In those systems, spamming gets you seen by everyone.In recommender systems, spamming only impacts the fraction of the users who are in the immediate neighborhood and see the spammy recommendations. The payoff is much reduced and so is the incentive to spam.For more on that, please see my Jul 2006 post, "Combating web spam with personalization" and my Jan 2007 post, "SEO and personalized search".The same authors published a similar but much shorter article, "Attacks and Remedies in Collaborative Recommendation", in the May/June IEEE Intelligent Systems, but there is no full text copy of that article easily available online.

http://glinden.blogspot.com/2007/07/attacking-recommender-systems.html