Comparing State-of-the-Art Collaborative Filtering Systems

Candillier, Laurent; Meyer, Frank; Boullé, Marc

doi:10.1007/978-3-540-73499-4_41

Cited by 138 publications

(105 citation statements)

References 18 publications

(17 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The recommendation is generated according to heuristic or probabilistic methods [16], which are applied to the known ratings in order to predict an unknown rating of a user. Regarding image recommendation, there are useful methodologies and tools which can bring considerable benefits, since they allow users to keep a record of their ratings.…”

Section: Image Recommendation Typesmentioning

confidence: 99%

“…On the first place, a collaborative filtering algorithm has been implemented with the best configuration for our use case [16]: the user-based collaborative filtering. Therefore, Pearson correlation (cosine of deviation from the mean) has been used for obtaining the similarities between users (16), and the final predictions have been computed using a weighted sum of deviations from the mean (17): Where v ui is the rating of user u on item i, S u is the set of items rated by user u, and v¡¡ is the mean rating of user u.…”

Section: Improvement Unrmentioning

confidence: 99%

See 1 more Smart Citation

Social and Content Hybrid Image Recommender System for Mobile Social Networks

et al. 2012

View full text Add to dashboard Cite

One of the advantages of social networks is the possibility to socialize and personalize the content created or shared by the users. In mobile social networks, where the devices have limited capabilities in terms of screen size and computing power, Multimedia Recommender Systems help to present the most relevant content to the users, depending on their tastes, relationships and profile. Previous recommender systems are not able to cope with the uncertainty of automated tagging and are knowledge domain dependant. In addition, the instantiation of a recommender in this domain should cope with problems arising from the collaborative filtering inherent nature (cold start, banana problem, large number of users to run, etc.). The solution presented in this paper addresses the abovementioned problems by proposing a hybrid image recommender system, which combines collaborative filtering (social techniques) with content-based techniques, leaving the user the liberty to give these processes a personal weight. It takes into account aesthetics and the formal characteristics of the images to overcome the problems of current techniques, improving the performance of existing systems to create a mobile social networks recommender with a high degree of adaptation to any kind of user.

show abstract

Section: Image Recommendation Typesmentioning

confidence: 99%

Section: Improvement Unrmentioning

confidence: 99%

Social and Content Hybrid Image Recommender System for Mobile Social Networks

et al. 2012

View full text Add to dashboard Cite

show abstract

“…In the online phase, the prediction is as follows: [5], incremental training is achieved by using new ratings to update the average parameters (r kl ,ru,r k ,ri,r l ) in equation (3). However, new users or items are not assigned to clusters during the online phase.…”

Section: Incremental Cf Via Co-clusteringmentioning

confidence: 99%

“…Collaborative filtering (CF) uses purchase or rating information to recommend items based on similarity [3]. If two users have liked (or disliked) similar items up to now, it is likely that they will have the same behavior in the future.…”

Section: Introductionmentioning

confidence: 99%

Incremental collaborative filtering via evolutionary co-clustering

Khoshneshin

Street

2010

Proceedings of the Fourth ACM Conference on Recommender Systems

View full text Add to dashboard Cite

Collaborative filtering is a popular approach for building recommender systems. Current collaborative filtering algorithms are accurate but also computationally expensive, and so are best in static off-line settings. It is desirable to include the new data in a collaborative filtering model in an online manner, requiring a model that can be incrementally updated efficiently. Incremental collaborative filtering via co-clustering has been shown to be a very scalable approach for this purpose. However, locally optimized co-clustering solutions via current fast iterative algorithms give poor accuracy. We propose an evolutionary co-clustering method that improves predictive performance while maintaining the scalability of co-clustering in the online phase.

show abstract

“…But looking at the recommendation techniques used on such web sites shows there is room for innovation. Candillier et al (2007) presents an overview of recommendation techniques. These techniques are either based on internet users notations or content descriptions (userand item-based techniques using collaborative filtering), or based on matching Internet user profiles and content descriptions (content filtering), or based on hybrid techniques combining both approaches.…”

Section: Introductionmentioning

confidence: 99%

Automating Opinion Analysis in Film Reviews: The Case of Statistic Versus Linguistic Approach

Poirier

Bothorel

Neef

et al. 2011

Affective Computing and Sentiment Analysis

Self Cite

View full text Add to dashboard Cite

Community sites are by nature dedicated places to express and publish opinions. www.flixster.com is an example of participative web site, with dozens of millions of enthusiasts sharing their feelings/views on movies, providing positive feedback as well as vivid critics. For anyone interested in understanding net user expectations, such web sites are of major importance because they offer the opportunity to probe huge volume of user generated contents. But to actually benefit from those large amount of data, one has to be able to automatically extract users opinions. This is the challenge we tackle in this paper. Our goal is to exploit the various reviews written by a user in order to compute a model which can then be used to predict the user's verdict on a movie. We explore two different methods to extract opinions. The first one relies on a machine learning technique based on a naive bayesian classifier. The second method consists in applying NLP techniques to process opinions and build dictionaries : those dictionaries are then used to determine the polarity of a comment given the words it may contain. We did apply those two approaches to contents from flixster.com : the results we provide enable us to discern the most appropriate approach for a given set of data.

show abstract

Comparing State-of-the-Art Collaborative Filtering Systems

Cited by 138 publications

References 18 publications

Social and Content Hybrid Image Recommender System for Mobile Social Networks

Social and Content Hybrid Image Recommender System for Mobile Social Networks

Incremental collaborative filtering via evolutionary co-clustering

Automating Opinion Analysis in Film Reviews: The Case of Statistic Versus Linguistic Approach

Contact Info

Product

Resources

About