Though research on the Semantic Web has progressed at a steady pace, its promise has yet to be realized. One major difficulty is that, by its very nature, the Semantic Web is a large, uncensored system to which anyone may contribute. This raises the question of how much credence to give each source. We cannot expect each user to know the trustworthiness of each source, nor would we want to assign top-down or global credibility values due to the subjective nature of trust. We tackle this problem by employing a web of trust, in which each user maintains trusts in a small number of other users. We then compose these trusts into trust values for all other users. The result of our computation is not an agglomerate "trustworthiness" of each user. Instead, each user receives a personalized set of trusts, which may vary widely from person to person. We define properties for combination functions which merge such trusts, and define a class of functions for which merging may be done locally while maintaining these properties. We give examples of specific functions and apply them to data from Epinions and our BibServ bibliography server. Experiments confirm that the methods are robust to noise, and do not put unreasonable expectations on users. We hope that these methods will help move the Semantic Web closer to fulfilling its promise.

Trust management for the Semantic Web

Providing ambient media services in the pervasive environments is a challenging issue. This is due to the fact that users have different satisfaction level in using different media services in varying contexts. We address this issue by proposing a gain-based media service selection mechanism. Gain refers to the extent a media service is satisfying to a user in a particular context. In our proposed mechanism, the gain is dynamically computed by adopting a user-centered approach that includes user's context, profile, interaction history, and the reputation of a service. The dynamically computed gain is used in conjunction with the cost of using a service (e.g. media subscription and energy consumption cost) to derive our service selection mechanism. We adopt a combination of greedy and dynamic programming based solution to obtain a set of services that would maximize the user's overall gain in the ambient environment by minimizing the cost constraint. Experimental results demonstrate the potential of this approach.

Gain-based selection of ambient media services in pervasive environments

This study integrates cognitive and affective attitudes as the primary factors of influence to formulate a Ubiquitous Computing Acceptance Model (UCAM), which is intended to predict whether potential users will accept ubiquitous computing (u-computing). The goal of this study was to empirically assess the UCAM by identifying the factors that contribute to user attitudes and intentions in u-computing environments. This study is one of the first to empirically evaluate the relationships among trust, security, privacy, usefulness, ease of use and behavioural intention in the ubiquitous city context.

Ubiquitous Computing Acceptance Model: end user concern about security, privacy and risk

In crowdsourcing systems, the interests of contributing participants and system stakeholders are often not fully aligned. Participants seek to learn, be entertained, and perform easy tasks, which offer them instant gratification; system stakeholders want users to complete more difficult tasks, which bring higher value to the crowdsourced application. We directly address this problem by presenting techniques that optimize the crowdsourcing process by jointly maximizing the user longevity in the system and the true value that the system derives from user participation. We first present models that predict the "survival probability" of a user at any given moment, that is, the probability that a user will proceed to the next task offered by the system. We then leverage this survival model to dynamically decide what task to assign and what motivating goals to present to the user. This allows us to jointly optimize for the short term (getting difficult tasks done) and for the long term (keeping users engaged for longer periods of time). We show that dynamically assigning tasks significantly increases the value of a crowdsourcing system. In an extensive empirical evaluation, we observed that our task allocation strategy increases the amount of information collected by up to 117.8%. We also explore the utility of motivating users with goals. We demonstrate that setting specific, static goals can be highly detrimental to the long-term user participation, as the completion of a goal (e.g., earning a badge) is also a common drop-off point for many users. We show that setting the goals dynamically, in conjunction with judicious allocation of tasks, increases the amount of information collected by the crowdsourcing system by up to 249%, compared to the existing baselines that use fixed objectives.

/pdf/getting-more-for-less-optimized-crowdsourcing-with-dynamic-4ol6lptlny.pdf

Getting More for Less: Optimized Crowdsourcing with Dynamic Tasks and Goals

As the promise of the Internet of Things (IoT) materializes in our everyday lives, we are often challenged with a number of concerns regarding the efficacy of the current data privacy solutions that support the pervasive components at play in IoT. The privacy and security concerns surrounding IoT often manifests themselves as a treat to end-user adoption and negatively impacts trust among end-users in these solutions. In this paper, we present a reference software architecture for building cloud-enabled IoT applications in support of collaborative pervasive systems aimed at achieving trustworthiness among end-users in IoT scenarios. We present a case study that leverages this reference architecture to protect sensitive user data in an IoT application implementation and evaluate the response of an end-user study accomplished through a survey.

A Reference Architecture for Improving Security and Privacy in Internet of Things Applications

Current pervasive environments should contain mechanisms, such as personalization, that adapt the environment to help the user meet their individual needs. However, manually creating, maintaining and utilizing a preference set is no easy task for a user, requiring continued time and effort. A more desirable approach is to implicitly build and maintain the preference set by using monitoring and learning mechanisms and apply such preferences when required on behalf of the user. This paper introduces the Daidalos Personalization and Learning system which monitors user behaviour and context to not only build and maintain dynamic preferences but also to apply them in a dynamic fashion. An example scenario is presented to demonstrate how such mechanisms are used to adapt a pervasive environment on a user?s behalf.

/pdf/adapting-pervasive-environments-through-machine-learning-and-655wlgv9b3.pdf

Adapting Pervasive Environments through Machine Learning and Dynamic Personalization

The importance of personalization in many application areas is now well accepted and this is true of both telecommunications and web services. One important research area which spans telecommunications and web services is that of pervasive systems. Here the role of personalization is essential in order for a pervasive system to function effectively without the need for unnecessary interaction with the user and for improving the user experience by providing appropriate support for the actions taken by the user. However, in order to provide personalization in such a system the system needs to build up an adequate set of user preferences. This is a non-trivial task. This paper describes briefly the approaches to personalization included in the development of the pervasive system in the Daidalos project. It discusses some of the strategies involved in managing user preferences and the approaches being taken in the Daidalos system. In particular in the second phase a novel approach is proposed to deal with this problem.

/pdf/managing-user-preferences-for-personalization-in-a-pervasive-4ajsqw9du8.pdf

Managing User Preferences for Personalization in a Pervasive Service Environment

There are various critical privacy issues that need to be addressed in the majority of smart space environments. This paper elaborates on the design of a privacy protection framework for Personal Self-Improving Smart Spaces (PSSs), a concept introduced by the Persist project Consortium. Compared to other smart spaces, such as smart homes and vehicles, this new paradigm provides a truly ubiquitous and fully personalisable user-centric environment. However, the information that needs to be collected, processed and distributed in such an environment is by nature highly privacy-sensitive, as it includes user profile data and preferences, as well as data regarding the past, current and even future user activities and context in general. In this respect, the designed privacy framework aims to address all privacy issues that arise by providing facilities which support multiple digital identities of PSS owners and privacy preferences for deriving privacy policies based on the context and the trustworthiness of the third parties that interact with PSSs.

A Privacy Framework for Personal Self-Improving Smart Spaces

Pervasive computing envisions a world where people are
surrounded by numerous communication and computing interconnected devices that are invisible and assist users in their everyday tasks in a seamless unobtrusive manner. Most pervasive computing research initiatives aim towards the realization of smart spaces, i.e. fixed spaces that provide pervasive features in a static and geographically limited environment. To bridge these isolated pervasive spaces,
the EU project Persist has introduced the concept of self-improving Personal Smart Spaces (PSSs) that follow their owners wherever they go. This paper provides an overview of the context and preference learning facilities that have been designed to support the realization of PSSs and enhance their proactivity and self-improvement features.

A Hybrid Preference Learning and Context RefinementArchitecture

A wireless pervasive computing environment needs to dynamically adapt its functionality and behaviour to changes in the resources and services available to a user at any point in time. For a mobile user this means that as the user changes location, the pervasive environment should take account of any changes to the services that are available. Thus initially when the user requests a service the most appropriate service must be provided, but as the user moves around, the service provided could change. To handle this, a wireless pervasive system needs to provide dynamic service composition (and re-composition) based on the user's personal preferences and current context. The Daidalos project is developing a platform to support pervasive services, which provides dynamic personalized service composition. This paper describes the problems and the role of personalization in the approaches adopted in Daidalos to deal with dynamic service composition and re-composition. The basic ideas have been prototyped and demonstrated, and are currently being integrated and extended.

/pdf/personalized-dynamic-composition-of-services-and-resources-1ue1c8ycpb.pdf

Sarah McBurney

Papers

Adapting Pervasive Environments through Machine Learning and Dynamic Personalization

Managing User Preferences for Personalization in a Pervasive Service Environment

A Privacy Framework for Personal Self-Improving Smart Spaces

A Hybrid Preference Learning and Context RefinementArchitecture

Personalized Dynamic Composition of Services and Resources in a Wireless Pervasive Computing Environment