Stephen Barrett

Extracting trust from domain analysis: a case study on the wikipedia project

The problem of identifying trustworthy information on the World Wide Web is becoming increasingly acute as new tools such as wikis and blogs simplify and democratize publications. Wikipedia is the most extraordinary example of this phenomenon and, although a few mechanisms have been put in place to improve contributions quality, trust in Wikipedia content quality has been seriously questioned. We thought that a deeper understanding of what in general defines high-standard and expertise in domains related to Wikipedia – i.e. content quality in a collaborative environment – mapped onto Wikipedia elements would lead to a complete set of mechanisms to sustain trust in Wikipedia context. Our evaluation, conducted on about 8,000 articles representing 65% of the overall Wikipedia editing activity, shows that the new trust evidence that we extracted from Wikipedia allows us to transparently and automatically compute trust values to isolate articles of great or low quality.

Computational Trust in Web Content Quality: A Comparative Evalutation on the Wikipedia Project

Computational Trust in Web Content Quality: A Comparative Evalutation on the Wikipedia Project

Pedagogy and Processes for a Computer Programming Outreach Workshop—The Bridge to College Model

This paper describes a model for computer programming outreach workshops aimed at second-level students (ages 15-16). Participants engage in a series of programming activities based on the Scratch visual programming language, and a very strong group-based pedagogy is followed. Participants are not required to have any prior programming experience. An empirical evaluation was undertaken to evaluate to what extent the model was successful in: (1) giving the participants a deeper understanding of what studying a computing degree and working in the computing profession entails; and (2) increasing their interest in pursuing a third-level qualification in a computer-related area.

A Translation Mechanism for Recommendations

An important class of distributed Trust-based solutions is based on the information sharing. A basic requirement of such systems is the ability of participating agents to effectively communicate, receiving and sending messages that can be interpreted correctly. Unfortunately, in open systems it is not possible to postulate a common agreement about the representation of a rating, its semantic meaning and cognitive and computational mechanisms behind a trust-rating formation. Social scientists agree to consider unqualified trust values not transferable, but a more pragmatic approach would conclude that qualified trust judgments are worth being transferred as far as decisions taken considering others’ opinion are better than the ones taken in isolation. In this paper we investigate the problem of trust transferability in open distributed environments, proposing a translation mechanism able to make information exchanged from one agent to another more accurate and useful. Our strategy implies that the parties involved disclose some elements of their trust models in order to understand how compatible the two systems are. This degree of compatibility is used to weight exchanged trust judgements. If agents are not compatible enough, transmitted values can be discarded. We define a complete simulation environment where agents are modelled with characteristics that may differ. We show how agents’ differences deteriorate the value of recommendations so that agents obtain better predictions on their own. We then show how different translation mechanisms based on the degree of compatibility improve drastically the quality of recommendations. Distributed System Group Department of Computer Science and Statistics Trinity College Dublin, Dublin 2, e-mail: dondiop@cs.tcd.ie e-mail: longol@cs.tcd.ie e-mail: stephen.barrett@cs.tcd.ie

Temporal factors to evaluate trustworthiness of virtual identities

In this paper we investigate how temporal factors (i.e. factors computed by considering only the time-distribution of interactions) can be used as an evidence of an entity’s trustworthiness. While reputation and direct experience are the two most widely used sources of trust in applications, we believe that new sources of evidence and new applications should be investigated [1]. Moreover, while these two classical techniques are based on evaluating the outcomes of interactions (direct or indirect), temporal factors are based on quantitative analysis, representing an alternative way of assessing trust. Our presumption is that, even with this limited information, temporal factors could be a plausible evidence of trust that might be aggregated with more traditional sources. After defining our formal model of four main temporal factors - activity, presence, regularity, frequency, we performed an evaluation over the Wikipedia project, considering more than 12000 users and 94000 articles. Our encouraging results show how, based solely on temporal factors, plausible trust decisions can be achieved.

Presumptive selection of trust evidence

1. This paper proposes a generic method for identifying elements in a domain that can be used as trust evidences. As an alternative to external infrastructured approaches based on certificates or user recommendations we propose a computation based on evidences gathered directly from application elements that have been recognized to have a trust meaning. However, when the selection of evidences is done using a dedicated infrastructure or users collaboration it remains a well-bounded problem. Instead, when evidences must be selected directly from domain activity selection is generally unsystematic and subjective, typically resulting in an unbounded problem. To address these issues, our paper proposes a general methodology for selecting trust evidences among elements of the domain under analysis. The method uses presumptive reasoning combined with a human-based and intuitive notion of Trust. Using the method the problem of evidence selection becomes the critical analysis of identified evidences plausibility against the situation and their logical consistency. We present an evaluation, in the context of the Wikipedia project, in which trust predictions based on evidences identified by our method are compared to a computation based on domain-specific expertise.

P2P live video streaming in WebRTC

In this paper, we analyse the feasibility of implementing live video streaming protocols into web applications with the use of WebRTC. As a result of demand the distribution of video content requires ever increasing bandwidth. Although, specialised programs exist to distribute video content efficiently, web pages have up until recently not been able to leverage these technologies. WebRTC could serve as a solution by enabling peer-to-peer communication directly between browsers without any need for a server as an intermediary. The feasibility analysisis accompanied by a practical implementation of a peer-to-peer streaming protocol in WebRTC that runs natively in all browsers and an identification of optimal settings for such a protocol. Our work highlights current limitations and future challenges when implementing sophisticated peer-to-peer solutions using a technology that is still in its infancy. Finally, we provide preliminary experimental data on WebRTC which measures the performance of such a system in a laboratory environment.

Self-adaptive application for indoor wayfinding for individuals with cognitive impairments

This article focuses on describing a Model Driven Architecture (COCA-MDA) approach that facilitates the development of self-adaptive application for indoor wayfinding for individuals with cognitive impairments. COCA-MDA provides the following benefits: 1) It enables the architecture to anticipate several behavioural variations based on the context and the specific needs of the individuals with cognitive impairments. 2) It enables the application to proactively anticipate or reactively address unforeseen changes through support by a dynamic-decision making and policy framework. The policy framework is based on a stable description of software models and proprieties. 3) It can decompose the application into several architectural units to allow developers to decide which part of the architecture should be notified when a specific context condition occurs.

Cognitive effort for multi-agent systems

Cognitive Effort is a multi-faceted phenomenon that has suffered from an imperfect understanding, an informal use in everyday life and numerous definitions. This paper attempts to clarify the concept, along with some of the main influencing factors, by presenting a possible heuristic formalism intended to be implemented as a computational concept, and therefore be embedded in an artificial agent capable of cognitive effort-based decision support. Its applicability in the domain of Artificial Intelligence and Multi-Agent Systems is discussed. The technical challenge of this contribution is to start an active discussion towards the formalisation of Cognitive Effort and its application in AI.

The Emergence of Useful Bias in Self-focusing Genetic Programming for Software Optimisation

The use of Genetic Programming GP to optimise increasingly large software code has been enabled through biasing the application of GP operators to code areas relevant to the optimisation of interest. As previous approaches have used various forms of static bias applied before the application of GP, we show the emergence of bias learned within the GP process itself which improves solution finding probability in a similar way. As this variant technique is sensitive to the evolutionary lineage, we argue that it may more accurately provide bias in programs which have undergone heavier modification and thus find solutions addressing more complex issues.