3,462 Hits in 5.5 sec

Improving Neighborhood-Based Collaborative Filtering by Reducing Hubness

Peter Knees, Dominik Schnitzer, Arthur Flexer
2014 Proceedings of International Conference on Multimedia Retrieval - ICMR '14  
In traditional memory-based (or neighborhood-based) recommenders, this is accomplished by, first, selecting a number of similar users (or items) and, second, combining their ratings into a single user's  ...  For recommending multimedia items, collaborative filtering (CF) denotes the technique of automatically predicting a user's rating or preference for an item by exploiting item preferences of a (large) group  ...  INTRODUCTION Personalized recommendation of media items to users by means of mining and exploiting preference data of many other users, also known as collaborative filtering (CF) [8] , has become a central  ... 
doi:10.1145/2578726.2578747 dblp:conf/mir/KneesSF14 fatcat:lycg7z22ergtxfhllhsx7foy3e

Exploiting Popularity and Similarity for Link Recommendation in Twitter Networks

Jun Zou, Faramarz Fekri
2014 ACM Conference on Recommender Systems  
The second approach adapts the collaborative filtering algorithms to incorporate popularity in addition to similarity.  ...  The first approach employs the rank aggregation technique to combine rankings generated by popularity-based and similarity-based recommendation algorithms.  ...  ACKNOWLEDGMENTS This material is based upon work supported by the National Science Foundation under Grant No. IIS-1115199.  ... 
dblp:conf/recsys/ZouF14 fatcat:4o7hqnauivcttjag7pyzs656dq

Improving Neighborhood-Based Collaborative Filtering by a Heuristic Approach and an Adjusted Similarity Measure

Yasser El Madani El Alami, El Habib Nfaoui, Omar El Beqqali
2015 International Conference on Big Data Cloud and Applications  
Collaborative filtering" is the most used approach in recommendation systems since it provides good predictions.  ...  This paper presents a new algorithm for neighborhood selection based on two heuristic approaches.  ...  Neighbor-based approach is mainly divided in two analogous categories: users-based collaborative filtering [15] and itembased collaborative filtering [20] .  ... 
dblp:conf/bdca/AlamiNB15 fatcat:i5pob2cc2rdtrbuxu57eawlxhq

On over-specialization and concentration bias of recommendations

Panagiotis Adamopoulos, Alexander Tuzhilin
2014 Proceedings of the 8th ACM Conference on Recommender systems - RecSys '14  
Focusing on the problems of over-specialization and concentration bias, this paper presents a novel probabilistic method for recommending items in the neighborhood-based collaborative filtering framework  ...  This performance improvement is in accordance with ensemble learning theory and the phenomenon of "hubness" in recommender systems.  ...  Common recommenders, such as collaborative filtering (CF) algorithms, recommend products based on prior sales and ratings.  ... 
doi:10.1145/2645710.2645752 dblp:conf/recsys/AdamopoulosT14a fatcat:pf5l7ysi3jdcde77v3lra5xtpu

A Link Analysis Approach to Recommendation under Sparse Data

Zan Huang, Daniel Zeng, Hsinchun Chen
2004 Americas Conference on Information Systems  
Collaborative filtering is one most successful approach to recommendation reported in the literature.  ...  Under this consumer-product graph, we propose to explore the global graph structure to facilitate collaborative filtering under sparse data.  ...  ACKNOWLEDGMENTS This research was supported in part by the following grants: NSF Digital Library Initiative-II, "High-performance Digital Library Systems: From Information Retrieval to Knowledge Management  ... 
dblp:conf/amcis/HuangZC04 fatcat:vhvrvkby6vdrtlyo6ptel73s3a

How does high dimensionality affect collaborative filtering?

Alexandros Nanopoulos, Miloš Radovanović, Mirjana Ivanović
2009 Proceedings of the third ACM conference on Recommender systems - RecSys '09  
A crucial operation in memory-based collaborative filtering (CF) is determining nearest neighbors (NNs) of users/items.  ...  The first is similarity concentration and the second is the appearance of hubs (i.e. points which appear in k-NN lists of many other points).  ...  INTRODUCTION Memory-based collaborative filtering (CF) is a successful recommender system technology that produces recommendations by matching user preferences to those of other users.  ... 
doi:10.1145/1639714.1639771 dblp:conf/recsys/NanopoulosRI09 fatcat:j6o54f4fvje33dqagrqn54jrzq

A comprehensive empirical comparison of hubness reduction in high-dimensional spaces

Roman Feldbauer, Arthur Flexer
2018 Knowledge and Information Systems  
Several hubness reduction methods based on different paradigms have previously been developed.  ...  Many machine learning algorithms rely on nearest neighbor search and some form of measuring distances, which are both impaired by high hubness.  ...  Acknowledgements Open access funding provided by Austrian Science Fund (FWF).  ... 
doi:10.1007/s10115-018-1205-y pmid:32647403 pmcid:PMC7327987 fatcat:zd4agqoxovdx7icr7opz3losca

Can Shared Nearest Neighbors Reduce Hubness in High-Dimensional Spaces?

Arthur Flexer, Dominik Schnitzer
2013 2013 IEEE 13th International Conference on Data Mining Workshops  
SNN is shown to reduce hubness but less than other approaches and, contrary to its competitors, it is only able to improve classification accuracy for half of the data sets.  ...  Computation of secondary distances inspired by shared nearest neighbor (SNN) approaches has been shown to reduce hubness and concentration and there already exists some work on direct application of SNN  ...  The existence of the hub problem has also been reported for music recommendation based on collaborative filtering instead of on audio content analysis [18] .  ... 
doi:10.1109/icdmw.2013.101 dblp:conf/icdm/FlexerS13 fatcat:pgbotxa5dvhwlo4juwhi7shpom

Peer-to-peer based recommendations for mobile commerce

Amund Tveit
2001 Proceedings of the 1st international workshop on Mobile commerce - WMC '01  
In this paper a Peer-to-Peer (P2P) based collaborative filtering architecture for the support of product and service recommendations for mobile customers is considered.  ...  Mobile customers are represented by software assistant agents that act like peers in the processing of recommendations.  ...  This work is partially supported by the Norwegian Research Council in the framework of the Distributed Information Technology Systems (DITS) program and the ElCo-mAg project.  ... 
doi:10.1145/381461.381466 dblp:conf/wmc/Tveit01 fatcat:ppfnzjuz4fdutoprkyg3ehh22m

A co-opetitive framework for the hub location problems in transportation networks

Rahimeh Neamatian Monemi, Shahin Gelareh, Saïd Hanafi, Nelson Maculan
2017 Optimization  
In our matheuristic algorithm, the neighborhood solutions are evaluated using a Lagrangian relaxation-based approach.  ...  The LSPs would like to cooperate with each other by establishing joint edges with limited capacities connecting their service networks.  ...  The authors are thankful of the anonymous referees for their constructive comments that significantly improved the article.  ... 
doi:10.1080/02331934.2017.1295045 fatcat:zawvvqr4snfpfndwyrsophlcsi

Hubness-aware kNN classification of high-dimensional data in presence of label noise

Nenad Tomašev, Krisztian Buza
2015 Neurocomputing  
We evaluate the potential impact of hub-centered noise by defining a hubness-proportional random label noise model that is shown to induce a significantly higher kNN misclassification rate than the uniform  ...  Hubness is an important aspect of the curse of dimensionality that has a negative effect on many types of similarity-based learning methods.  ...  According to Figure 11 , an increase in neighborhood size improves kNN classification performance on these particular datasets, as it reduces the influence of noise.  ... 
doi:10.1016/j.neucom.2014.10.084 fatcat:433silukgbhaxlwifzzfpnghya

CoBaR: Confidence-Based Recommender [article]

Fernando S. Aguiar Neto, Arthur F. da Costa, Marcelo G. Manzato
2018 arXiv   pre-print
Neighborhood-based collaborative filtering algorithms usually adopt a fixed neighborhood size for every user or item, although groups of users or items may have different lengths depending on users' preferences  ...  In this paper, we propose an extension to a non-personalized recommender based on confidence intervals and hierarchical clustering to generate groups of users with optimal sizes.  ...  Some works [1, 5] studied the effects of neighborhood size in collaborative filtering, but usually the length of clusters is not personalized for each user.  ... 
arXiv:1808.07089v1 fatcat:khocgpt7ejemrmw2itxw6lla24

Choosing ℓp norms in high-dimensional spaces based on hub analysis

Arthur Flexer, Dominik Schnitzer
2015 Neurocomputing  
Hub objects have a small distance to an exceptionally large number of data points while anti-hubs lie far from all other data points.  ...  The hubness phenomenon is a recently discovered aspect of the curse of dimensionality.  ...  Acknowledgements This research was supported by the Austrian Science Fund (FWF, Project P27082).  ... 
doi:10.1016/j.neucom.2014.11.084 pmid:26640321 pmcid:PMC4567076 fatcat:g5oxavf7svgm5jv6zkyq6sv2da

A modified fuzzy C-means algorithm for collaborative filtering

Jinlong Wu, Tiejun Li
2008 Proceedings of the 2nd KDD Workshop on Large-Scale Recommender Systems and the Netflix Prize Competition - NETFLIX '08  
filtering (CBF) and collaborative filtering (CF).  ...  Table 1 : Artist recommendation network properties for collaborative filtering (CF), content-based audio filtering (CB), and (AMG) expert-based.  ... 
doi:10.1145/1722149.1722151 fatcat:br7cxd6ljfhmpl2kdtqvi43q6e

Hubness in Unsupervised Outlier Detection Techniques for High Dimensional Data –A Survey

R.Lakshmi Devi, R. Amalraj
2015 International Journal of Computer Applications Technology and Research  
high dimensional data and role of hubness.  ...  Hubness is an aspect for the increase of dimensionality pertaining to nearest neighbors which has come to an attention.  ...  Also proves that this new algorithm improves the computational complexity with reduced number of iterations.  ... 
doi:10.7753/ijcatr0411.1004 fatcat:l3lxmvua2je7xpvxnp7k7ni2pa
« Previous Showing results 1 — 15 out of 3,462 results