Given the vast amount of online information covering almost all aspects of human endeavor, the Internet, especially the Web, is clearly a fertile ground for data mining research from which to extract valuable knowledge. Web mining is the application of data mining techniques to extract knowledge from Web data, including Web documents, Web hyperlink structure, and Web usage logs.Traditional Web mining research has been mainly focused on addressing the information overload problem. Many information retrieval (IR) and artificial intelligence (AI) techniques have been adopted or developed to identify relevant information from the Web to meet users' specific information needs. However, most existing studies do not fully explore the social and behavioral aspects of the Web. Thus, the primary goal of this dissertation is to develop an integrated research framework that extends traditional Web mining methodologies to fully explore the technical, social, and behavioral aspects of Web knowledge discovery.My dissertation framework is composed of technical and social/behavioral components. In the technical component of my dissertation, a set of domain specific Web collection building, Web content and link structure mining, and Web knowledge presentation techniques were developed. These techniques were tested in a series of case studies to demonstrate their effectiveness and efficiency in facilitating knowledge discovery in various domains.The social/behavioral component of my dissertation is to explore the application of Web mining technology as a new means to study the social interactions and behavior of Web content providers and users. Several case studies were conducted to extract knowledge on covert organizations' resource allocation plans, information management policies, and technical sophistication using Web mining techniques. Such knowledge would be very difficult to obtain through other means.The major contributions of this dissertation are twofold. First, it proposed a set of new Web mining techniques that can help facilitate knowledge discovery in various domains. Second, it demonstrated the effectiveness and efficiency of applying Web mining techniques in extracting social and behavioral knowledge in different contexts.

Huge growth of the Internet is due to the large number of websites and web services through which information is easily accessible. E-commerce and e-services obtain much private data from users for various reasons such as advertising, marketing, etc. Collection, storage, and usage of private data are subject to various standards, privacy laws, and regulations. To adhere to these legal requirements, many privacy services, such as secure data transmission, authentication, notice, and consent, are required. Inclusion of these required privacy services early in the life cycle of the software development is preferred and advocated, but not fully adhered to. Inclusion of privacy services in legacy software and currently developed software is required. We describe software architecture and a system for automatic inclusion of privacy services, under the supervision of privacy expert, into web pages after the development phase of the Software Development Life Cycle. This will help organizations to adhere to standards, privacy laws, and regulations when collecting private data online from its clients. We also describe a prototype that we have developed as a proof-of-concept to demonstrate the feasibility of our approach.

<p> Modern Internet services are multi-tiered and are typically hosted in virtualized shared platforms. While facilitating flexible service deployment, multi-tier architecture introduces significant challenges for Quality of Service (QoS) provisioning in hosted Internet services. Complex inter-tier dependencies and dynamic bottleneck tier shift are challenges inherent to tiered architectures. Hard-to-predict and bursty session-based Internet workloads further magnify this complexity. Virtualization of shared platforms adds yet another layer of complication in managing the hosted multi-tier Internet services. </p><p> We consider three critical aspects of Internet service management for improved performance and quality of service provisioning : admission control, dynamic resource provisioning and service differentiation. This thesis concentrates on statistical learning based approaches for multi-tier Internet service management to achieve efficient, balanced and scalable services. Statistical learning techniques are capable of solving complex dynamic problems through learning and adaptation with no <i>priori</i> domain-specific knowledge. We explore the effectiveness of supervised and unsupervised learning in managing multi-tier Internet services. </p><p> First, we develop a session based admission control strategy to improve session throughput of multi- tier Internet services. Using a supervised bayesian network, it achieves coordination among multiple tiers resulting in a balanced service. Second, we promote session-slowdown, a novel session-oriented metric for user perceived performance. We develop a regression based dynamic resource provisioning strategy, which utilizes a combination of offline training and online monitoring, for session slowdown guarantees in multi-tier systems. Third, we develop a reinforcement learning based coordinated combination of admission control and adaptive resource management for multi-tier Internet service differentiation and performance improvement in a shared virtualized platform. It addresses limitations of supervised learning by integrating model-independence of reinforcement learning and self-learning of neural networks for system scalability and agility. Finally, we develop an user interface based Monitoring and Management Console, intended for an administrator to monitor and fine tune the performance of hosted multi-tier Internet services. </p><p> We evaluate the developed management approaches using an e-commerce simulator and an implementation testbed on a virtualized blade server system hosting multi-tier RUBiS benchmark applications. Results demonstrate the effectiveness and efficiency of statistical learning approaches for QoS provisioning and performance improvement in virtualized multi-tier Internet services.</p>

<p> Interest in virtual worlds has grown within academia and popular culture. Virtual worlds are persistent, technologically-mediated, social spaces. Academic literature focuses on issues such as identity, sociality, economics, and governance. However studies of governance focus on internal or external modes of control; less attention has been paid to institutions of governance that operate within both the virtual and real worlds. </p><p> In EVE Online, the Council of Stellar Management (CSM) represents a joint venture between developers and users to shape the direction of EVE's virtual society. As a group of elected representatives, the CSM represents societal interests to the game's developer, Crowd Control Productions (CCP). The CSM structures the relationship between CCP and the player base, and shapes how these institutions manage the development process. At the same time, cultural and political conventions of EVE's players at times work against these structures as CCP and the CSM seek to attend to their own interests. </p><p> In this thesis, I examine the intersection of culture, power, and governance, and illustrate the consequences these negotiations of power have for the inhabitants of EVE Online. The historical circumstances that led to the CSM's creation shape its reception among the community. As a model of governance, the CSM was designed as a deliberative democracy to generate community consensus. This feedback is channeled to developers through elected representatives. However, these channels of information hindered discussions necessary for true democracy. I examine how power is generated, leveraged, and mediated by the two cultures in which the CSM is embedded: EVE and Icelandic cultures. I also illustrate the authority and legitimacy of the CSM from the standpoint of its constituents. The CSM is understood within the same cultural frameworks as in-game power structures. </p><p> Primary research was carried out during a one-year period in 2012. During this time, I joined SKULL SQUADRON, a large corporation with a neutral diplomatic mission. Snowball sampling was used to find informants. Three main methods were used to interview participants: face-to-face interviews, text-based interviews through EVE's communication channels, and voice interviews conducted over Skype, an internet-based communications program.</p>

<p> As the popularity of Web 2.0 grows, the relationship between the users generating the content of the site and the groups and companies that own these sites is coming into focus. While in previous years, users often were passive users of websites, now they are actively involved in the sites, providing the content that is consumed. This creates a relationship that can be fraught with conflict as all involved have differing ideas of how these sites should function.</p><p> By analyzing three incidents in the history of LiveJournal, an online blogging and social communication site, this thesis explored how these communities of users and the organizations that own these sites interact. The information for this analysis was gathered through online participant observation, survey, and systematic archival mining, covering the time period from the founding of the site in 1999 until early 2012. I analyzed how the term "community" is operationalized by these stakeholders, how these communities formed and functioned, and how ideas of ownership impacted these interactions and relationships. By understanding these issues, companies and communities can find ways to build partnerships to sustain and improve their sites rather than being locked in ongoing conflict.</p>

<p> People are increasingly looking to online social communities as ways of communicating. However, even as participation in social networking is increasing, online communities often fail to coalesce. Noted success factors for online communities are linked to the community's purpose and culture. They are also related to structures that allow for increased volume of exchange and quality of conversation. Ravelry.com provides a case of a successful, large-scale, online community that has information exchange and conversation as its foci. These activities are supported through the work of thousands of volunteer moderators who sometimes manage groups with more than 3 million members. However, little is known about organizing and supporting volunteer groups to allow for such large-scale growth. </p><p> To find information on moderators' roles and tasks, and how they are supported in the Ravelry community, a study was conducted in 2 sequential phases. Phase I consisted of a survey of 73 moderators who led large, active groups. Phase II consisted of interviews with 8 moderators who led different types of groups within Ravelry, having purposes that range from purely social conversation to technical forums on craft-related work. </p><p> Findings indicated that the tasks moderators performed did not vary greatly, despite differences in their group's purpose and culture. Common among most moderators' duties were encouraging group participation through stimulating discussion or organizing craft-based activities, resolving conflict between group members, and maintaining their site through routine housekeeping tasks. Moderators are motivated to volunteer to do these tasks by love of their group members, and are united by a common interest in their craft. Moderators are most often supported by informal networks of moderators in their own or in similar groups. </p><p> These findings give insight into how to structure large, asynchronous, online conversation-based groups, and how to define a role for people to manage them. It provides an understanding of the work that moderators do, and how their work allows a sense of place to be established for informal learning. </p>

No description available.

<p> Microblogging service Twitter has taken the world by storm since its inception in 2006, growing from 340,000 users in July 2007 to 500 million active users in March 2013. At the same time, Twitter and other social media platforms are opening up new possibilities for organizations to engage with and be responsive to their stakeholders and to the public in general.</p><p> Despite the widespread use of social media among nonprofit organizations, very little empirical evidence is available concerning publics' responses to the messages they are sent. This thesis describes how stakeholders respond to different communication practices on Twitter. Focusing on the organizations on the <i>"Nonprofit Times 100"</i> list of 2011 as well as the list of nonprofit organizations with the most Twitter followers, this study combines qualitative and quantitative analyses at both the message level and the organizational level to develop an initial understanding of effective Twitter practices among nonprofit organizations.</p><p> Until recently, nonprofits have not fully taken advantage of the interactive possibilities Twitter has the potential to provide. After analyzing 3,415 tweets by 50 nonprofits, it became clear that it was difficult, if not impossible, to identify the "perfect" tweet - the type of tweet that would be most likely to elicit all types of engagement. This study's results suggest that nonprofits should target specific tweets toward retweeting and favoriting engagement, and others toward conversations - and not expect the same tweet to achieve both.</p>

The goal of the thesis presented herein is to provide an overview of current issues in auditory display design and to suggest how these issues may be applied to the development of applications for the World Wide Web (WWW). The software developed as part of this thesis, the SprocketPlug plug-in for Netscape Navigator, provides a tool for exploring various auditory display techniques at three levels of WWW development: HTML, Javascript, and Java. / The strength of SprocketPlug is that it enables WWW developers to incorporate interactive spatialized sound as an integrated component of WWW documents and applications. The implementation of SprocketPlug is based on currently available technology: the Netscape plug-in architecture, Netscape LiveConnect, and the Apple SoundSprocket.

<p> The purpose of this study is to specifically look into one businesses website and online efforts to engage potential new hires in providing them with valuable company information. The research questions focused on how effective the company's online advertisements were, as well as the interaction and involvement of the online advertisements. The researcher of this study utilized communication theories based on Todd Gitlin's media torrent theory of oversaturation and Joseph Walther's social information processing theory to conduct research. To conduct the research a survey was given to 20 prospective job seekers and five face-to-face interviews were conducted. The participation in the survey was anonymous and no tracking information was kept. The interviewees that participated in the interview were current employees that were only asked about their application process and the role that social media played in their online job search. The study concludes that most of the prospective job seekers and current employees find social media platforms to be an easy way to find open positions because these online advertisements direct the applicant to the company website for further information. However, many employees also preferred to learn about the company face-to-face, rather than online. </p>