Collaborative recommendation: A robustness analysis

By Michael P. O'Mahony, Neil J. Hurley, Nicholas Kushmerick, Guenole C. M. Silvestre. 2004

376 citations recorded by the source corpus.

Subfields: Computer Science Applications, Information Systems, Management Science and Operations Research

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Science Applications

1. J. Konstan
2. John Riedl
3. Michael D. Ekstrand
4. Neil Zhenqiang Gong
5. Bamshad Mobasher
6. Shyong K. Lam
7. Harald Steck
8. Sarasi Lalithsena

In Information Systems

1. Bamshad Mobasher
2. J. Konstan
3. John Riedl
4. Michael D. Ekstrand
5. Charu C. Aggarwal
6. Gediminas Adomavicius
7. Jonathan L. Herlocker
8. Robin D. Burke

In Management Science and Operations Research

1. Bamshad Mobasher
2. Robin D. Burke
3. J. Konstan
4. John Riedl
5. Jonathan L. Herlocker
6. Shilad Sen
7. Taghi M. Khoshgoftaar
8. Xiaoyuan Su

Citations in the corpus

• Poisoning Attacks to Graph-Based Recommender Systems (2018, background): “ecurity A‡acks. „ese a−acks aim to spoof a recommender system such that a target item is recommended to as many or few users as possible. Speci•cally, poisoning a‹acks (also known as shilling a‹acks) [19, 21, 24] aim to inject fake users with fake rating scores to the system such that a bad recommen”
• Shilling attacks against recommender systems: a comprehensive survey (2012, methodology): “In O’Mahony et al. (2004c), the authors scrutinize the ability of a CF system to make predictions despite noisy ratings.”
• Data Mining for Web Personalization (2007, background): “Power of an attack [ 96 ] on the other hand measures the average change in the gap between the predicted and target rating for the target item.”
• Trust-Aware Collaborative Filtering for Recommender Systems (2004, background): “However, subverting standard CF techniques is very easy [10].”
• Effective Attack Models for Shilling Item-Based Collaborative Filtering Systems (2005, methodology): “In [6] two evaluation measures were introduced: robustness and stability.”
• Recommender Systems (2016, neutral): “Robustness and stability metrics for recommender systems are discussed in [158, 329, 393, 444].”
• Shilling recommender systems for fun and profit (2004, methodology): “One way in which the ACF algorithms we test are different from those used in the past [5, 14] is that we do not use negative correlations between items.”
• Collaborative Filtering Recommender Systems (2011, background): “For example, recommenders based on Pearson correlation without similarity damping are susceptible to shilling attacks with small attack profiles, as two users with few ratings in common have abnormally high Pearson correlations [108].”
• Collaborative Filtering Recommender Systems (2007, background): “O’Mahoney et al [42] showed that users could, in fact, artificially raise and lower predicting ratings.”
• Preventing shilling attacks in online recommender systems (2005, background): “While there is a lot of work in the field of developing collaborative filtering algorithms, only recently some papers have concentrated on developing shilling attack models [9, 12] and on benchmarking the robustness of recommender systems against shilling attacks [4, 8].”
• A Survey of Collaborative Filtering Techniques (2009, background): “contributed to solving the shilling attacks problem by analyzing robustness, a recommender system’s resilience to potentially malicious perturbations in the customer/product rating matrix [50].”
• Evaluating Recommendation Systems (2011, background): “Robustness (see Chapter 25) is the stability of the recommendation in the presence of fake information [45], typically inserted on purpose in order to influence the recommendations.”