| | |

inDiscover - NRC technology helps Canadian independent music industry

Music lovers now have access to technology that will automatically put them in touch with the independent Canadian music artists’ songs that they are likely to enjoy. inDiscover™ is a web portal powered by RACOFI Composer™, technology developed at NRC Institute for Information Technology (NRC-IIT), designed to help bring independent music to new listeners by recommending songs based on users’ tastes. The portal also offers independent Canadian music artists access to an effective distribution channel and a wider audience base. In this capacity, the technology has the potential to help resolve a critical issue for an important Canadian industry; illegal downloading of copyright-protected digital entertainment (music, video, or text).

"Indiscover.net is a central part of the Sympatico/MSN music strategy. It is the technology platform that we are using to reach out to the Canadian music community, to generate interest and awareness in independent musicians, and to lead in the development of new business models for music. Working with NRC, we've added new functionality and features that allow us to host and stream artists' music, and allow users to build play-lists and share them with friends."

Veronica Holmes is Senior Director, Broadband Innovation Bell Sympatico/MSN

First-of-its-kind system makes tailored, ‘word-of-mouth’ recommendations automatically

Before internet shopping, customers’ choices were typically limited by physical factors. Stores only had so much shelf and storage space, and therefore would only carry those popular products that would guarantee the best return on investment. With the advent of e-commerce, online companies are no longer restricted by physical barriers. Consumers are now faced with an overwhelming variety of product from which to choose.

‘Word of mouth’ is a common means of filtering available choices to come up with a meaningful list. Recommendations made by like-minded individuals (i.e. friends) can help consumers refine the list of available options ultimately make reaching a decision easier. But how can these filtering systems be used online?

Typical internet search engines return results that often do not capture the interest of the user and the person is therefore left to sift through hundreds of indexed web pages. As well, the results of searches are only as good as the search parameters that are used to generate them. If you don’t know exactly what you are looking for, the chances are that Google will not be able to find it for you.

In response to this overwhelming issue, NRC Institute for Information Technology (NRC-IIT) has developed RACOFI Composer™, a technology for building high-precision, adaptable recommender systems.

RACOFI Composer™ is a first-of-its-kind system with the core idea to leverage as much as possible what is known about a user as well as other similar users to make tailored, ‘word-of-mouth’ recommendations automatically. Essentially, the personalized recommendations are based on the evaluation of thousands of similar users and thousands of similar criteria simultaneously which returns meaningful search results within seconds.

This is achieved by combining collaborative filters, rule-based recommender systems and a unique mathematical algorithm so that the system draws information from both data and inferences.

inDiscover, used for rating and recommending music, is the first commercial web application of RACOFI Composer technology and is licensed to Bell Canada, one of Canada’s largest corporations. Used by Bell on their MSN/Sympatico entertainment website, InDiscover.net is a web portal, powered by NRC’s RACOFI Composer™, to promote independent Canadian musicians by providing personalizing song recommendations, in the form of play-lists, to users.

By using NRC’s technology, inDiscover.net is able to predict how probable a given user would like a specific song based on how the user and other people similar to the user previously rated other songs and therefore, provide suggested play-lists that are meaningful. Users also rate songs according to different categories (eg. good work-out song?) and therefore, the site is able to compose personalized play-lists for each category (e.g., a play-list of suggested work-out songs) for users. The site implements the strategy “I rate, therefore I get personalized recommendations.”

The performance of inDiscover.net, using the power of NRC’s RACOFI Composer™, is remarkable; recommendations can be culled from collection of a million rated songs in a matter of seconds. And as of March 2005 the web portal was attracting over 20,000 hits per week. InDiscover.net presented the 2007 Independent Music Awards as well as the Favorite Artist/Band website of the year award.

Through inDiscover.net, NRC’s technology promises to make a substantial impact and contribution to the solution of a critical issue for an important Canadian industry; illegal downloading of copyright-protected digital entertainment (music, video, or text) by empowering the industry with a legitimate alternative to illegal downloading. As well, it offers independent Canadian musicians access to an effective distribution channel and a wider audience base which will increase this market segment.

NRC has received both development and licensing fees for inDiscover and stands to receive royalties as the site become commercially viable. As well, NRC benefits from a partnership with a major Canadian Telecom company with potential for future developments.

inDiscover researchers at NRC will continue creating an information-rich platform for people and machines to collaboratively explore and define the Canadian independent music landscape as the website combines state-of-the-art machine learning technology with the latest social networking techniques. NRC’s technology also represents promising enabling technology with other commercial and non-commercial applications across a wide range of market segments.