This project aims to describe how workflow systems can integrate and make use of context information from context rich environments, to enhance the execution of workflow processes. Context information can for example be used to control transitions between activities, activity enactment and process/activity coordination. A dynamic contextual environment also requires that a workflow system is capable of responding to contextual events. A set of requirements for a context-aware workflow system, based on existing workflow standards, theory behind context-aware computing and activity theory, will be presented and elaborated. Prototypes, which illustrate how these requirements can be implemented in a standard based workflow system, are also provided. Based on the solutions presented in the prototypes, a new interface for a workflow enactment service is presented. This new interface serves as the link between the contextual environment and the workflow system. We also present a solution for handling context related exception states. The definition of basic terms in workflow systems are expanded to better support context-aware behaviour. Ideas and solutions for more complex requirements not met in our prototypes are also discussed, such as situated activity coordination.

Business processes are increasingly used by organizations to automate their activities. Written in languages like WS-BPEL, they allow an institution to describe precisely its internal operations. As the pace of change increases, however, both organizations and their internal processes are required to be more flexible; they have to account for an increasing amount of externally-driven environment state, or context, and modify their behavior accordingly. This puts a significant burden on business-process programmers, who now have to source, track, and update context from multiple entities, in addition to implementing and maintaining core business logic. Implementing this state-maintenance logic in a WS-BPEL business process is involved. This is because WS-BPEL business processes are modeled as if they were the only thing operating in, and making changes to, the business environment. This mental model does not reflect the real world, where organizations and entities depend on state that is outside their control – state that is modified independent of, and concurrent with, the organization’s activities. This makes it hard for business-process programmers to write context-dependent processes in a concise manner. This thesis presents a solution to this problem based on the notion of a context variable for WS-BPEL business processes. It describes how context variables are designed using the WS-BPEL language-extension mechanism, and how these variables can be used in business processes. It also outlines an architecture for offering context in the web services environment that uses constructs from the WS-Resource Framework specification. It shows how changes in context can be propagated from these context sources to WS-BPEL context variables using WS-Notification-based publish/subscribe. The design also includes a standards-compliant method for extending web-service responses with references to context sources. Finally, a prototype validating the overall system is described, and enhancements for increasing the utility of context variables proposed. This solution offers significant advantages: it builds on established practices and well-understood message-exchange patterns, leverages widely used languages, frameworks and specifications, is standards compliant, and has a low barrier-to-entry for business-process programmers. Moreover, when compared to existing alternatives, this solution requires significantly less process logic and fewer interface changes to maintain constantly changing environment state.

Business processes are increasingly used by organizations to automate their activities. Written in languages like WS-BPEL, they allow an institution to describe precisely its internal operations. As the pace of change increases, however, both organizations and their internal processes are required to be more flexible; they have to account for an increasing amount of externally-driven environment state, or context, and modify their behavior accordingly. This puts a significant burden on business-process programmers, who now have to source, track, and update context from multiple entities, in addition to implementing and maintaining core business logic. Implementing this state-maintenance logic in a WS-BPEL business process is involved. This is because WS-BPEL business processes are modeled as if they were the only thing operating in, and making changes to, the business environment. This mental model does not reflect the real world, where organizations and entities depend on state that is outside their control – state that is modified independent of, and concurrent with, the organization’s activities. This makes it hard for business-process programmers to write context-dependent processes in a concise manner. This thesis presents a solution to this problem based on the notion of a context variable for WS-BPEL business processes. It describes how context variables are designed using the WS-BPEL language-extension mechanism, and how these variables can be used in business processes. It also outlines an architecture for offering context in the web services environment that uses constructs from the WS-Resource Framework specification. It shows how changes in context can be propagated from these context sources to WS-BPEL context variables using WS-Notification-based publish/subscribe. The design also includes a standards-compliant method for extending web-service responses with references to context sources. Finally, a prototype validating the overall system is described, and enhancements for increasing the utility of context variables proposed. This solution offers significant advantages: it builds on established practices and well-understood message-exchange patterns, leverages widely used languages, frameworks and specifications, is standards compliant, and has a low barrier-to-entry for business-process programmers. Moreover, when compared to existing alternatives, this solution requires significantly less process logic and fewer interface changes to maintain constantly changing environment state.

Some agent-based systems depend on eliciting ratings from the user. However, the user’s willingness to provide ratings is limited due to requisite demands of attention and effort. From a human-centered view, we redefine providing ratings as expressing interest. We develop a new interface component for parameter setting, the In-Context Slider, which reduces physical effort and demand on attention by using fluid mouse gestures and in-context interaction. We hypothesize that such an interface should make interest expression easier for the user. We evaluated the In-Context Slider as an interest expression component compared with a more typical interface. Participants performed faster with the In-Context Slider. They found it easier to use and more natural for expressing interest. We then integrated the In-Context Slider in the agent-based system, combinFormation. We compared the In-Context Slider with combinFormation’s previous interest expression interface. Of the participants that effectively used both interfaces, most expressed more interest with the In-Context Slider. Participants’ experience reports described the In-Context Slider as easier to use while developing collections to answer open-ended information discovery questions. This research is relevant for many applications in which users provide ratings, such as recommender systems, as well as for others in which values need to be adjusted on many objects that are concurrently displayed.

This study looks into the field of intercultural miscommunication and misunderstandings on small multicultural Non Profit Organization (NPO) seen from the expats point of view. It is a study based on a qualitative method-design, including a micro-ethnographic study and qualitative interviews on a small NPO on Bali, Indonesia called Project Clean Uluwatu (PCU) that contain volunteers from all over the world. These misunderstandings and miscommunications that occurred on PCU was mostly between local people born on Bali and foreign people working on PCU, due to many reasons, starting with the rapid pace of globalization and that culture don’t evolve in the same speed. Plausible explanations for this, that are raised in this thesis, is that people make sense of events in different ways, especially if the individuals within a misunderstanding belong to both a high context culture and a low context culture and aby that communicate in different ways. This thesis also explain them through Karl Wieck’s sensemaking perspective by applying 7 properties that describes how individuals make sense out of miscommunication by perceiving the event in different ways due to individuals former knowledge, their intentions and own identity.

Thesis (Ph.D.)--Bowling Green State University, 2005. / Document formatted into pages; contains ix, 139 p. Includes bibliographical references.

Ethnic identity is an important aspect of individuals’ sense of self. For individuals identified as ethnic minorities, ethnic identity has been found to be a potential protective factor for overall well-being. Multiracials (i.e., individuals identified with two or more races) are one of the fastest growing minority populations in the United States. Limited research examining multiracials’ ethnic identity development currently exists. Furthermore, there is a paucity of ethnic identity literature examining longitudinal ethnic identity growth from adolescence to emerging adulthood. Ethnic minority adolescents, such as multiracials, and emerging adults are often at higher risk for lower psychological well-being and higher substance use. Therefore, understanding developmental trajectories and factors that contribute to ethnic identity development allows for clinicians to work with ethnic minority individuals in ways that are empowering and facilitate success. The current study utilizes Hierarchical Linear Modeling (HLM) to examine longitudinal growth trajectories of ethnic identity among multiracial and monoracial groups (White, Black, Latino/a, and Other [includes Asian, Pacific Islander, Native American, and Other]). HLM was also used to examine the relationships between social factors (i.e., experiences of discrimination, teasing by peers, and bullying) and school contextual factors (i.e., school climate, school safety, and perception of teachers) factors in the development of ethnic identity over time. The present study drew from an ethnically diverse sample of individuals living in the Pacific Northwest who were assessed each year from grade 6 to 9 and once during emerging adulthood (N = 593). Results indicated small linear increases of ethnic identity over time. In general, ethnic identity increased from Grades 6 through 8, decreased from Grade 8 to 9, and increased again from Grade 9 to emerging adulthood. Multiracials’ ethnic identity growth, however, did not differ from the identity growth of individuals within monoracial groups. Negative peer interactions significantly contributed to decreases in ethnic identity scores for individuals from Grade 8 to 9. School context did not significantly contribute to changes in ethnic identity growth. Findings suggest that individuals’ ethnic identity changes over time, and is significantly impacted by peer interactions.

Made available in DSpace on 2014-06-12T15:50:47Z (GMT). No. of bitstreams: 2 arquivo1988_1.pdf: 1433897 bytes, checksum: 2bd67eddaeadba13aa380ec5c913b7e0 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2008 / Uma das propostas da Web Semântica é fornecer uma grande diversidade de serviços de diferentes domínios na Web. Estes serviços são, em sua maioria, colaborativos, cujas tarefas se baseiam em processos de tomada de decisão. Estas decisões, por sua vez, serão mais bem embasadas se considerarem a maior quantidade possível de informação relacionada às tarefas em execução. Neste sentido, este cenário encoraja o desenvolvimento de técnicas e ferramentas orientadas para a integração de informação, procurando soluções para a heterogeneidade das fontes de dados. A arquitetura baseada em mediação, utilizada no desenvolvimento de sistemas de integração de informações tem como objetivo isolar o usuário das fontes de dados distribuídas utilizando uma camada intermediária de software chamada de mediador. O mediador, em um sistema de integração de informações, utiliza um esquema global para a execução das consultas do usuário que são reformuladas em sub-consultas de acordo com os esquemas locais das fontes de dados. Neste caso, um processo de integração de esquemas gera o esquema global (esquema de mediação) como resultado da integração dos esquemas individuais das fontes de dados. O problema maior em integração de esquemas é a heterogeneidade das fontes de dados locais. Neste sentido, a resolução semântica é primordial. A utilização de métodos puramente estruturais e sintáticos na integração de esquemas é pouco eficaz se antes não houver a identificação do real significado dos elementos dos esquemas. Um processo de integração de esquemas tem como resultado um esquema global integrado e um conjunto de mapeamentos inter-esquema e usualmente, compreende algumas etapas básicas como: pré-integração, comparação, mapeamento e unificação de esquemas e geração do esquema de mediação. Em integração de esquemas, resolução de nomes é o processo que determina a qual entidade do mundo real um dado elemento de esquema se refere, levando em consideração um conjunto de informações semânticas disponíveis. A informação semântica necessária para resolução de nomes, em geral, é obtida de vocabulários genéricos e/ou específicos de um determinado domínio de conhecimento. Nomes de elementos podem apresentar significados diferentes dependendo do contexto semântico ao qual eles estão relacionados. Assim, o uso de informação contextual, além da de domínio, pode trazer uma maior precisão na interpretação dos elementos permitindo modificar o seu significado de acordo com um dado contexto. Este trabalho propõe uma abordagem de resolução de nomes baseada em contexto para integração de esquemas. Um de seus pontos fortes é a utilização e modelagem da informação contextual necessária à resolução de nomes em diferentes etapas do processo de integração de esquemas. A informação contextual está modelada utilizando uma ontologia, o que favorece a utilização de mecanismos de inferência, compartilhamento e reuso da informação. Além disto, este trabalho propõe um processo de integração de esquemas simples e extensível de forma que seu desenvolvimento se concentrasse principalmente nos requisitos relacionados à resolução de nomes. Este processo foi desenvolvido para um sistema de integração de informações baseado em mediação, que adota a abordagem GAV e XML como modelo comum para intercâmbio de dados e integração de fontes de dados na Web

As the development of context aware applications has evolved, there has been a corresponding increase in need for more sophisticated system. The aim for this thesis is the development of a dynamical P2P network system which is based on locations. The P2P network is self organizing and in a lightweight format. Modern technical solutions including   AGPS have facilitated the work associated with the ability to position users and modern mathematical solutions such as spherical trigonometry provides the P2P system with the necessary accuracy even for short distances. The P2P system works on different Java platforms including JSE, JME and Android. Unfortunately the 3G network distributor has not yet solved a NAT traversal problem, which means that the P2P network self organization and architecture has only been proved by means of simulations. Another problem is that certain mathematical formulas are required for the spherical trigonometry and the limitation for JME is that it is unable to handle inverse trigonometry. However, the Android and JSE versions can form a correct P2P network, under the condition that the Android device uses a WIFI connection point outside the 3G distributor network system. This thesis reports the successful testing of the locations-based P2P network. / MediaSense

In general, recommender systems attempt to estimate user preference based on historical data. A context-aware recommender system attempts to generate better recommendations using contextual information. However, generating recommendations for specific contexts has been challenging because of the difficulties in using contextual information to enhance the capabilities of recommender systems. Several methods have been used to incorporate contextual information into traditional recommendation algorithms. These methods focus on incorporating contextual information to improve general recommendations for users rather than identifying the different context applicable to the user and providing recommendations geared towards those specific contexts. In this thesis, we explore different context-aware recommendation techniques and present our context-aware coupled matrix factorization methods that use matrix factorization for estimating user preference and features in a specific contextual condition. We develop two methods: the first method attaches user preference across multiple contextual conditions, making the assumption that user preference remains the same, but the suitability of items differs across different contextual conditions; i.e., an item might not be suitable for certain conditions. The second method assumes that item suitability remains the same across different contextual conditions but user preference changes. We perform a number of experiments on the last.fm dataset to evaluate our methods. We also compared our work to other context-aware recommendation approaches. Our results show that grouping ratings by context and jointly factorizing with common factors improves prediction accuracy.