Dynamic Personalized Recommendation on iTV for Seniors


    Dynamic personalized recommendation of information on public and social services on iTV for seniors: a case study


    Academic & Research Project


    Specific Audiences




    Luiz David Campelo






    The dissemination and adequate access to information about Services of General Interest are constitutional rights of the citizens and play a major role in structuring a more egalitarian society based on the democratization of knowledge. However, despite the increasing amount of information available and the evolution of information and communication technologies (ICT), senior citizens, often characterized by lower levels of digital literacy and info-inclusion, often struggle to access information about policies and services that they can benefit from. With specific informational needs and free time due to retirement, seniors tend to use TV as a primary mean of information and entertainment. In this way, benefiting from the familiarity of these citizens with the TV, many innovative technological solutions have been leveraged this device. However, solely designing and employing technologically advanced features is not enough. It is necessary to develop personalized solutions to better adapt to seniors’ preferences and limitations. In this case, this concerns identifying which information is more appropriate to be provided for each senior. For example, information on health campaigns and social tariffs discounts should be tailored according to the user’s specific preferences and contextual factors (e.g. location and dates). That said, this research proposes a personalization strategy for the delivery of highvalued informative contents about Services of General Interest for the senior population. To this end, this work aims to leverage the informative videos exhibition through the integration of a Context-Aware Recommender System (CARS). The investigation was divided into three distinct phases, in a participatory design approach, so that the CARS is adequate to the specifics of this population segment, considering seniors’ opinions and indications in all phases of the study. In the first phase, data of the trinomial [Item x User x Context] is characterized. In addition, this phase was carried out with the collaboration of specialists in the areas of gerontology, public services, interactive TV and software engineering, as well as the collaboration of seniors recruited under the +TV4E project, through the application of interviews, focus groups and guided tests. In the second phase, the CARS is proposed according to the Data Model and the interaction scheme obtained from the results of the previous phase. A hybrid filtering algorithm is proposed to generate the recommendations. Finally, in the third and last phase, a prototype was developed and integrated in the scope of + TV4E project, in order to validate the CARS, in a domestic environment, for a period of two weeks, and with the support of 21 senior residents in the district of Aveiro. The analysis of the results, based on user interactions and interviews, corroborate the usefulness and appropriateness of the personalization strategy proposed by CARS.


    Thesis (PT)

    Dynamic Personalized Recommendation on iTV for Seniors