The sociolinguistic concept of ‘community’ has undergone profound changes since it was first introduced in the middle of the twentieth century. Originally conceived to describe groups of individuals and families with shared traditions and a long history of living and working closely together (think of a village or an urban neighbourhood), the concept has now expanded to include a variety of ways in which people can connect with each other: attachments which are often longdistance, transitory, mobile, multiple. As we shall see, developments in sociolinguistic ideas about community to some extent parallel perceived changes in the real world, given the quickening pace of globalisation, the shake-up of traditional structures and sources of authority, the greater possibility for mobility, and people’s increasing ability to use digital communications to establish networks beyond their immediate vicinity. If we look at how …

The Isis toolkit offers the sophisticated capabilities required to analyze digital personas and provide investigators with clues to the identity of the individual or group hiding behind one or more personas.

The collection of representative corpus samples of both child language and online (CMC) language varieties is crucial for linguistic research that is motivated by applications to the protection of children online. In this paper, we present an extensive survey of corpora available for these two areas. Although a significant amount of research has been undertaken both on child language and on CMC language varieties, a much smaller number of datasets are made available as corpora. Especially lacking are corpora which match requirements for verifiable age and gender metadata, although some include self-reported information, which may be unreliable. Our survey highlights the lack of corpus data available for the intersecting area of child language in CMC environments. This lack of available corpus data is a significant drawback for those wishing to undertake replicable studies of child language and online language varieties.

The collection of representative corpus samples of both child language and online (CMC) language varieties is crucial for linguistic research that is motivated by applications to the protection of children online. In this paper, we present an extensive survey of corpora available for these two areas. Although a significant amount of research has been undertaken both on child language and on CMC language varieties, a much smaller number of datasets are made available as corpora. Especially lacking are corpora which match requirements for verifiable age and gender metadata, although some include self-reported information, which may be unreliable. Our survey highlights the lack of corpus data available for the intersecting area of child language in CMC environments. This lack of available corpus data is a significant drawback for those wishing to undertake replicable studies of child language and online language varieties.

Privacy control mechanisms for online social networks (OSNs) offer little by way of managing access to a user's personal information by third-party applications (TPAs). Most OSNs provide an "accept all or nothing" mechanism for managing permissions from TPAs to access a user's private data. In this paper, we propose an approach that makes all requests for private data from TPAs explicit and enables a user to exert fine-grained access control over what profile data can be accessed by individual applications. Equally importantly, our approach also allows users to share their access control configurations for TPAs with their friends who can reuse and rate such configurations. This is particularly beneficial to novice users or those new to a particular TPA or an OSN. We present an implementation of our approach for managing privacy for third-party Facebook applications and report an initial evaluation (N=50). A significant proportion of our sample (76%) found the collaborative privacy management approach useful in determining the type of applications one might use based on its privacy rankings and noted a raised awareness about data privacy issues arising from use of TPAs.

Online social networks are now common place in day-to-day lives. They are also increasingly used to drive social action initiatives, either led by government or communities themselves (e.g., SeeClickFix, LoveLewisham.org, mumsnet). However, such initiatives are mainly used for crowd sourcing community views or coordinating activities. With the changing global economic and political landscape, there is an ever pressing need to engage citizens on a large-scale, not only in consultations about systems that affect them, but also involve them directly in the design of these very systems. In this paper we present the UDesignIt platform that combines social media technologies with software engineering concepts to empower communities to discuss and extract high-level design features. It combines natural language processing, feature modelling and visual overlays in the form of “image clouds” to enable communities and software engineers alike to unlock the knowledge contained in the unstructured and unfiltered content of social media where people discuss social problems and their solutions. By automatically extracting key themes and presenting them in a structured and organised manner in near real-time, the approach drives a shift towards large-scale engagement of community stakeholders for system design.

Child protection social work's current concern remains largely with offline harm and injury and separate from the focus on child protection in relation to risks online and children's use of computer-mediated communication and the internet. This paper draws on original research reporting children's methods for assessing who they are talking to online, as part of a wider study designed to develop software tools for improved online child safety. It finds that children use similar methods to detect identity online as they do offline, reinforcing the interconnections between mediated and unmediated communication and an ontology of childhood in a digital world that is neither online nor offline, but both. Drawing on the concept of cyborg youth and understanding contemporary child protection in this context allows a re-visioning of the social work role that demands attention to computer-mediated literacy in the assessment of all vulnerable children and in the support of their parents and carers.

Peer-to-peer file sharing has become popular for many kinds of resource location and distribution applications including file sharing, distributed computation, multi-media messaging and content distribution. Peer-to-peer approaches also have significant potential for supporting large scale, decentralised service oriented computing. This chapter discusses each class of contemporary P2P architecture in turn and discusses the suitability of each architecture class for supporting service oriented computing. Future trends in peer-to-peer architectures are then discussed and multi-layer peer-to-peer architectures are highlighted as a promising platform for supporting service oriented computing. This chapter then concludes with a discussion of outstanding issues that must be addressed before peer-to-peer architectures can offer adequate support for service oriented computing.

Recent years have seen a rapid rise in the number and use of online social networks. Such social networks vary in nature from chat systems, for example, MSN, Skype and IRC, to online communities, such as, MySpace and YouTube, through to file sharing systems, for instance, peer-topeer networks: Gnutella, BitTorrent, FastTrack, etc. Amongst the many types of ‘risk’ on the internet as mentioned in the Byron review in the UK and Internet Safety Technical Task Force in the US, these social networks pose two significant risks in terms of child exploitation. The first major type of risk is paedophiles and other child sex offenders predating on children. Children actively participate in social interactions using forums such as chat rooms and web-based communities. Offenders can use such forums to predate on children, or even to plan the commission of sexual offences against children. These concerns are reflected by the formation of the Virtual Global Taskforce and specialist UK enforcement agencies and Scottish legislation to criminalise the ‘grooming’ of children in chat rooms in October 2004. The second risk is the offence of distributing and sharing child abuse media. Child sex offenders can formulate their own social networks using mechanisms, such as file-sharing systems, in order to distribute and share child abuse media. The scale of distribution of illegal media (including child abuse media) on such file-sharing systems was highlighted by a recent study at Lancaster University, which found that 1.6% of searches and 2.4% of responses on the Gnutella peer-to-peer network relate to illegal sexual content. Given the system’s scale, these results suggest that, on the Gnutella network alone, hundreds of searches for illegal images occur each second. The study also found that, of those users sharing illegal sexual content, 57% were solely devoted to such distribution while half of the material shared by another 17% involved such content. The EPSRC funded Isis project (www.comp.lancs.ac.uk/isis/) aims to help tackle these risks by developing an ethics-centred monitoring framework and toolkit that will support law enforcement agencies in the task of policing online social networks for the purpose of protecting children. Specifically it seeks to tackle three significant research challenges: 1. How to identify active child sex offenders across online communities? Distinguishing between “innocent” interaction amongst children or children and adults, and the predatory advances of paedophiles and other child sex offenders who will often masquerade as a child or friend in order to gain a potential victims trust 2. How to identify the core distributors of child abuse media? Accurately identifying child abuse media from the plethora of material that exists within file sharing systems; media that offenders often describe using specialised vocabulary that evolves and changes over time. 3. How to ensure that any technical solutions maintain ethical practices? Balancing the benefits of using technology for child protection with the need to protect innocent users from the potential of falsely being identified as child sex offenders and safeguarding their privacy. Given the vast amount of information that is communicated within online social networks, new monitoring and analysis technologies are required. Within Isis novel techniques are being developed that will support the analysis of chat logs and the non-invasive real-time monitoring of file sharing networks by drawing upon natural language processing practices. These can be used, for example, to help determine whether certain terminology repeatedly crops up, or whether users have trademark phrases that they use. Sophisticated statistical analysis techniques allow for the creation of language profiles for certain group types or even for an individual. These have already been successfully used to automatically distinguish between speech from males and females, from over 35 year olds and under 35 year olds, and also speakers from different regions within the UK. As part of Isis these techniques will be extended to allow the differentiation of child and adult language, and by extension the detection of adults masquerading as children. Law enforcement personnel will also be able to build language profiles of known paedophiles, which could then be used to assist in determining if they are re-offending. It should be noted, however, that our goal is not automation but to provide support for detecting potential sexual offences through analysis of large amounts of data which cannot manually be analysed in an efficient manner. In parallel the Isis project is also studying the ethical and social implications of the technology that is being developed and, as a result, further informing its development. To date there has been a lack of suitable case studies in the computer ethics literature and appropriate guidance for technology developers to incorporate ethical considerations within the development cycle. Within Isis we are also developing new understandings of user-centred development methods for highly sensitive systems. These, in turn, will benefit future related developments and help mitigate the effects of adverse outcomes that impact on public acceptability.

The growth of mobile devices with near PC equivalent capabilities has brought with it the possibility of mobile Peer-to-Peer (P2P) systems. However, the unique nature of mobile devices introduces new challenges that need to be considered during the development process, especially when considering critical aspects such as system security. This chapter presents the PEPERS Development Methodology (PDM), a tool-supported methodology that aims to assist designers in developing secure mobile P2P systems, and encourages them to consider specific mobile P2P design issues from an early stage. The PDM is demonstrated within the context of a real-world case study of a system developed for a security company.

In the public sector (particularly in the UK in light of recent reforms i.e. the Local Government Act 2000, etc.) a greater degree of accountability and public involvement or intervention has become the norm in public infrastructure projects, partially under the rubric of "stakeholder engagement". This paper seeks to discuss public involvement in a law-enforcement technology (Isis), which operates on a covert basis in the detection and prevention of child abuse activities across a number of social networking facilities. Our contribution to the development of Isis is to perform an ethics centered consultation process with stakeholders who will contribute to the design and deployment of the end software package. To that end, we have sought to develop a "Modified Participatory Design" approach, utilizing the knowledge gained from the HCI community with regards to more traditional design projects and adapting this body of work to questions of ethics, privacy, corporate and civic responsibility, monitoring and awareness issues, etc. in an effort to create a fluid and agile communication process between stakeholders and designers, thus taking account of the ethical issues around Isis as design occurs.

Peer-to-peer file sharing exploded into the public consciousness with the release of Napster in 1999. Napster sparked a revolution in computer-mediated communication and within a few years, peer-to-peer had replaced client-server as the dominant communications paradigm of the Internet. Peer-to-peer has also redefined the role of home users, empowering them to produce and distribute content free from control or censorship by third parties. The importance of the peer-to-peer revolution has been reflected by significant research and business activity, although considerable technical and socio-technical problems still remain. This chapter discusses the emergence of peer-to-peer systems as a medium for computer-mediated communication and how these systems have evolved to tackle new challenges. This is done in the context of a case-study of Gnutella, one of today’s most significant peer-to-peer file sharing systems.

Since the release of Napster in 1999, peer-to-peer file-sharing has enjoyed a dramatic rise in popularity. A 2000 study by Plonka on the University of Wisconsin campus network found that file-sharing accounted for a comparable volume of traffic to web applications, while a 2002 study by Saroiu et al. on the University of Washington campus network found that file-sharing accounted for more than treble the volume of web traffic observed, thus affirming the significance of P2P in the context of Internet traffic. Empirical studies of peer-to-peer traffic are essential for supporting the design of next-generation peer-to-peer systems, informing the provisioning of network infrastructure and underpinning the policing of peer-to-peer systems. This paper surveys existing work in the field of peer-to-peer monitoring and based upon this assessment of the state-of-the-art describes the design and implementation of the Open P2P tracing system, which aims to improve the research community’s understanding of P2P file sharing systems by providing continuous and up-to-date traffic data which is anonymized and made freely accessible to all interested parties. Data from this system has been used in a variety of projects and papers, which are used to illustrate the broad range of research that can benefit from an open tracing system.

This paper proposes a means to specify the semantics of fault tolerant Web services at an abstract level using semantics adapted from queuing system theory. A framework that supports the implementation of specified fault-tolerance is also described. Based on our work, we show how the redundancy and diversity characteristics of a service-oriented system can be expressed and implemented in a Web-service application.