Search

Over the last few years, we have witnessed a generalization of the consumption of video contents through different media (computers, television, mobile phones, tablets,…) that converge into a unified user experience in an (almost) all-IP technological environment. At the same time, there have been significant advances in the generation and characterization of multimedia contents, due to the advent of social networking and the consolidation of standards like moving picture experts group (MPEG)-4 and MPEG-7. Finally, it is becoming easier and more natural to interact with multimedia contents, thanks to devices like tactile screens and touchpads. Overall, this landscape enables great possibilities for advertising and e-commerce, far beyond the traditional business models of broadcast TV. In this paper, we present a novel realization of two non-invasive advertising and e-commerce models to exploit those possibilities, grounded on a cloud-based personalization engine. This system cannot only identify the most interesting products to advertise to each user at any time, but also compose personalized applications on the fly, ready to run on different consumer device platforms. Our experiments show that this proposal can be very positively perceived by the end users. [ABSTRACT FROM PUBLISHER]

Challenges today in the Web personalization and adaptation research area range from heterogeneous user needs and user environment issues such as current location and time to multi-channel delivery of applications concerning Web-based content related to services, educational multimedia, entertainment, commerce, etc. Nowadays, most Web personalization systems implement various techniques to extract the user profiles, which serve as the main component of such systems. Based on given user preferences and navigation behavior, the Web-based content returns the requested personalized result. The main scope of this paper is to present the various techniques employed by such systems with regard to extraction of user profiles and to introduce a comprehensive user profilethat includes user perceptual preference characteristics. It further analyzes the main intrinsic user characteristics such as visual, cognitive and emotional-processing parameters that are incorporated as well as the ‘traditional’ user profile characteristics that together tend to give the most optimized, adapted and personalized outcome. It finally presents a Web adaptation and personalization system that implements the proposed comprehensive user profile as well as the initial evaluation results. [ABSTRACT FROM PUBLISHER]

Scientific environments are known for being highly dynamic, subject to rapid evolution and demanding constant renewal and update. Additionally, science is a highly social arena. However, interpersonal collaboration and knowledge flow in scientific environments are usually more restricted. Collaboration is intense among small groups of people working on specific problems within a domain, but low between groups. As user profiling has been extensively used as a basis for recommendation, personalization and matchmaking systems, a better profile identification can improve interaction levels among researchers belonging to the same domain but working in different laboratories. Profiles may be constructed in two ways: either through explicit declaration by the user or through the observation of users’ actions. Many systems employ one approach to the exclusion of the other. We contend that a combined approach will yield better results, especially on scientific scenario, providing a mix of declared and inferred information. In this article, we present inference-based profiles in scientific environments (i-ProSE), an integrated system that dynamically creates and maintains scientific user profiles based both on declared information and on observed behaviour. The i-ProSE can be used to locate experts, deliver content, build communities, find collaborators for long-term projects or detect instantaneous opportunities for informal collaboration, which is presented with a short study case. [ABSTRACT FROM PUBLISHER]

For personal information agents, user profiles have to represent user interests and preferences in order to satisfy long-term information needs. An implicit assumption in user-profiling is the existence of persistent interests which, however, might suffer some changes over time. Each time the interests of a user change, his profile becomes inaccurate and the predictive quality decreases. Adaptation of user profiles is, therefore, an essential requirement for personal agents that need to be capable of adjusting their behavior quickly in order to shorten the period of reduced predictive quality. In this paper, a user-profiling technique named WebProfiler, which learns a hierarchical representation of user interests using conceptual clustering, is augmented with an adaptation strategy based on relevance feedback and time-based forgetting in order to deal with drifting interests. We empirically evaluate the performance of this strategy by analyzing its behavior on multiple scenarios of interest drifts and shifts. [ABSTRACT FROM PUBLISHER]

The article discusses various reports published within the issue, including one on Web personalization and adaptive hypermedia technologies, another on a user-profiling system called i-Prose, as well as a report on expert searching.