Preface

Information is today becoming a larger and steadily growing part of people’s lives. In the form of web-pages, magazines, music, movies and other television-material, regular people are slowly becoming mass-informationconsumers. This development comes with the problem of choice. In the pursuit towards content personalization, recommending systems have experienced great success and popularity bringing together ideas from areas like filtering techniques, machine learning, user profiling, information retrieval and much more [18]. Different methods have been used in recommending systems for calculating similarity knowledge with different results depending on the application domain. In this report these methods will be compared and a recommendation system prototype developed, optimized for digital television content. As another focus point mobileTV will also be considered as the second area of interest where the problem of choice becomes even more relevant as users’ tolerance becomes even harder to fulfill. Recommending systems come with some well known downfalls and the method of choice in each case depends on the content and the application domain in question. The objective with this report is to find the best suitable method for digital television content with the possibility of Video-on-Demand (VOD) service