Tired of wasting your time sifting through DVDs at the rental store only to end up with a mediocre film? Using a Movie Recommender System can help anyone interested in getting good movie recommendations.

However, it's not that simple as it looks. Each one of us has individual preferences. For instance, a certain product can please most of its buyers, but it doesn't satisfy all of them. When we analyze the people's taste in movies or songs, the individual preferences are more varied. A particular movie can be loved by a person and hated by another (even if it won an Oscar!).

To cope with this, there is a research domain called "Recommender Systems". Roughly speaking, such systems are software applications that are able to gather loads of data among preferences and behaviours of users/customers and use this data to suggest tailored items (products or services) to the customer.

One of the most popular techniques used to develop a "Recommender Systems" is the Collaborative Filtering approach (also called Social Filtering). This technique essentially automates the process of "word-of-mouth" recommendations. Items are recommended to an user based on the evaluations supplied by other users with similar preferences.

Nowadays, Recommender Systems are best known for their use in e-commerce Websites such as Amazon.com. In general, the more accurate the recommendations are, the bigger the sales will be. Hence, in e-commerce environments these systems can be a valuable CRM (Customer Relationship Management) tool by supporting customers in the decision-making and buying process.

Anyhow, Recommender Systems can be used to suggest anything: books, songs, restaurants and even jokes! Their usage to generate movie recommendations is not an innovative idea. The MovieLens project is considered one of the pioneers of this field, delivering movie recommendations since 1994. Currently, well-know companies like Blockbuster and Netflix have their own movie Recommender system also.

The Cinefilia was born in 2004 as part of the M.Sc. thesis of Leandro N. Ciuffo at Universidade Federal Fluminense (UFF) in Brazil. His thesis aimed at investigating the susceptibility of product raters on the WWW. So, the development of a recommender system itself was not the main target of his research.

Cinefilia on the Grid

Since September 2006 Leandro collaborates with the EELA project at the Intalian National Institute of Nuclear Physics (INFN) in Catania. Currently, Leandro has been assigned to work as Grid application manager, wich implies in coordinate and offer support to all EELA applications. In order to learn more about Grid technologies he decided to take his movie recommender system developed during his M.Sc. course and create a case study on the strategies adopted to port it on the Grid (see the latest technical paper about this work).

His work is also based on the fact that Collaborative Filtering (CF) relies on big datasets in order to generate accurate recommendations. However, for a large retailer like Amazon.com, with huge amount of data, tens of millions of customers and millions of distinct catalog items, generate accurate recommendations in real-time is impractical. Such limitation has driven the research of a vast amount of alternative solutions. Hence, several variations of the classic CF algorithm are present in the literature, but all of which reduce, in a certain extent, recommendation quality.

In this scenario where traditional CF systems have suffered from scalability issues, Grid computing appears as an innovative approach that can be used for running ever-larger workloads and services.

Grids first emerge within scientific communities, like High Energy Physics (HEP) and Bioinformatics. However, the enormous research activity in recent years has contributed to the development of new areas of interest. Commercial users have been attracted by this technology, which can potentially be utilized by industries and SMEs (Small and Medium-sized Enterprises) to offer new services with reduced costs and higher performance. Indeed, several international Grid projects such as EGEE, BEinGRID and Biz2Grid have been pushing industries and SMEs to have their business applications running in their Grid infrastructure.

This expansion from science to business is nearing Grids to "utility computing", where computing power is viewed as a utility, available on a pay-as-you-use basis, like gas or electricity. This is not yet the case, but there are several ongoing initiatives to develop new tools and Grid services that will allow the outsourcing of computing resources in the short term.

Hence, we understands that Grid computing technologies is getting even more mature and well known. As a consequence, new applications have been attracted to be deployed on it, even those that are not very CPU-intensive and that could be handled by small computing clusters.

Curriculum Cinefilia

Magazine article:
VENTON, D. For the love of movies: recommendations from the grid. International Science Grid This Week (iSGTW), 01/04/2009.

Full paper:
CIUFFO, L.N. Using Grids to Support Recommender Systems: A case study of generating movie recommendations on the EELA-2 infrastructure. Submitted to IBERGRID, 2009, Valencia (Spain).

Poster:
Using Grids to Support Recommender Systems: A case study of generating movie recommendations on the EELA-2 infrastructure. To be presented at IBERGRID, 2009, Valencia (Spain).

Presentation:
CIUFFO, L.N. Cinefilia Demo. Presented at EGEE User Forum, 2009, Catania (Italy).

Video: GridCast
Choose a movie using grid-powered recommendations. Presented at EGEE User Forum, 2009, Catania (Italy).

Short paper: (before porting Cinefilia to the Grid)
GARCIA, A.C.B.; CIUFFO, L.N. Applying the HYRIWYG incentive mechanism in a Recommender System. In: IEEE / WIC / ACM Internation Conference on Web Inteligence, 2005, Compiègne. Proceedings of the The 2005 IEEE/WIC/ACM International Conference on Web Intelligence (WI'05). Washington, DC, USA : IEEE Computer Society, 2005. v. 00. p. 770-773.

Acknowledgments

Just like "webifying" applications to run on a Web browser, Grid users need to "gridify" their applications to run on a Grid. This process may comprises the creation of additional bash scripts as well as changes in the original source codes in order properly interact with Grid services.

The "gridification" of Cinefilia has been carried out on the GILDA t-infrastructure prior to be ported on the EELA-2 infrastructure. The author would like to thank the GILDA team for all support provided.

Author also thanks the support provided by the EELA-2 Project. EELA-2 is co-funded by the European Commission in the frame of the Seventh EU Framework Programme - Research Infrastructures. This work reflects only the author’s views. The Community is not liable for any use that may be made of the information contained therein.