Orientador: Helen de Castro Silva / Banca: Mariângela Spotti Lopes Fujita / Banca: Ariadne Chloe Mary Furnival / Resumo: Com o desenvolvimento da Web e, conseqüentemente, das bibliotecas digitais, open archives, repositórios entre outros, novos sistemas e fontes de informação têm sido criados, colocando os usuários em um novo ambiente de busca e recuperação da informação, caracterizado pela sobreabundância de recursos informacionais dispostos no formato de hipertexto, aumentando significativamente as possibilidades de acesso. Diante deste ambiente digital de informação, uma das funções das Linguagens Documentárias é representar as informações dos documentos contidos nestes sistemas de informação, de tal forma que o usuário seja capaz de acessar essa informação para o uso, comunicação e, principalmente, para a geração de novos conhecimentos. No entanto, o problema-chave da Recuperação da Informação passa pela busca de procedimentos teóricos e metodológicos para a Organização e Representação da Informação, ou seja, uma das principais problemáticas é encontrar meios de aperfeiçoar os métodos de Organização da Informação ao nível conceitual, de forma a aumentar a acessibilidade pelos usuários finais. Diante disto, foi proposto um estudo exploratório de análise teórica de alguns modelos de comportamento de busca de informação que, de acordo com a literatura especializada, são os que têm maior impacto em termos de pesquisa na área. Tais modelos podem ser incorporados aos processos de tratamento temático da informação, permitindo a bibliotecários e indexadores a se darem conta das variáveis que interferem no processo de busca e recuperação da informação dos usuários e, desta forma, entenderem os fatores que afetam a organização e representação da informação. Como resultados considera-se que uma abordagem que procure colocar o usuário no centro das preocupações na Organização da Informação deve recorrer à experiência acumulada... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: With the development of the Web and, consequently, of the digital libraries, open archives, repositories between other, new systems and information sources have been created, placing the users in a new search and information retrieval environment, characterized by the abundance of informational resources disposed in the hypertext format, increasing the access possibilities significantly. Ahead of this information digital environment, one of the Indexing Languages functions is to represent the information of the documents contained in these information systems, so that the user be capable to access that information for the use, communication and, mainly, for the generation of new knowledge. However, the problem-key of the Information Retrieval goes by the search of theoretical and methodological proceedings for the Information Organization and Representation. In other words, the principal problematic is to find means of improving the methods of Information Organization at the conceptual level, in way to increase the accessibility for the end-users. Ahead this, an exploratory study of theoretical analysis was proposed of some information search behavior models that, in agreement with the specialized literature, they are the ones that have larger impact in research terms in the area. Such models can be incorporate to the information thematic treatment processes, allowing to librarians and indexers the if give bill of the intervening variables in the search and information retrieval process of the users and, this way, that they understand the factors that affect the information organization and representation. As results are considered that an approach that tries to place the user in the center of the concerns in the Information Organization should appeal to the accumulated experience in the calls studies of users. It is ended that the studies about the users' information behaviors... (Complete abstract click electronic access below) / Mestre

<p>With the rise of the Web, there is more interest among end-users to create different kinds of software that use elements from the Web or allow for networked interaction between users. Currently however, many available tools for this purpose are hard to use or lack a sufficient level of expressiveness. To provide new insights on the construction of tools that allow end-users to create their own Web-enabled software, this thesis explores design issues and consequences of applying the Turtle metaphor from the Logo-programming language to an end-user programming environment for Web-enabled applications. </p><p>In order to explore this, CrawLogo was created - a programmable end-user tool that supports the creation of Web-enabled applications using a Turtle-like control metaphor and language adapted from Logo. As a proof-of-concept, several Web-enabled applications were created using this new tool, including CrawLogo Pong, a somewhat alternative version of Atari’s classic Pong game, and a collaborative browsing environment, in which users can browse the Web together. The resulting CrawLogo environment allows for creating Web-enabled applications that - using more traditional programming languages - would be quite complex and require deep technical programming skills. Further, while utilizing a Turtle-like control metaphor in CrawLogo allows for the creation of some new types of applications and some new ways of interacting with the Web, it also raises new problems such as how to successfully design within the CrawLogo metaphor and how to create a meaningful representation of Turtle- geometry-based navigation on the Web.</p>

Abstract Following the vigorous development of the information and internet technology, the National Tax Administrations are now facing the challenges from the changes of the inner and outer environment. Due to the update of the information operating platform as well as the drive of e-government from the Executive Yuan, the government officials need to have abilities to learn new information technology; otherwise, they will fail to be qualified on their jobs. However, there are not too many studies about the degree of acceptance of officials on learning new technology. It is necessary to provide an approach to figure out acceptance degree of officials for information technology. We provide an integrated model called the Technology Acceptance Model on government officials. The model incorporates a wide variety of important factors into a theoretical framework provided by "Technology Acceptance Model, TAM" (Davis, 1989) and "Theory of Planned Behavior, TPB" (Ajzen, 1985). We collect data from a sample of staffs of the five local tax administrations and we analyze the data using the Structural Equation Modeling. The results of this research indicate that essential factors and show what factors affect the behavior of the government officials on accepting new technology and e-government. In addition, the model provides a good fit because of fit indices and explanatory powers for behavior. That is, we can use the model to predict the officials¡¦ behavior on using tax information system.

Att inventera är en påfrestande process där anställda samlar in data med en observationsundersökning. Ett exempel är livsmedelsbutiker som skriver ut meterlånga listor inför en lagerinventering där inventeraren anmärker skillnader på de fysiska varorna och vad som står på listorna. Inventering är något som de flesta företag gör men som skiljer sig från bransch till bransch. Eftersom det är ett resursslöseri att lösa samma problem flera gånger, utan att återanvända tidigare lösningar, har studenten valt att studera en relativt generell lösning på ett inventeringssystem som ska kunna användas av flera branscher.Syftet med denna studie är därför att studera vilka funktionalitetskrav som finns på ett inventeringssystem och vilka tekniska lösningar som kan användas vid utveckling av ett inventeringssystem.För att utreda detta tillämpades en abduktiv ansats där en fallstudie gjordes på inventering. En litteraturstudie gjordes på tekniska lösningar som kan användas vid utveckling av ett inventeringssystem.Studiens resultat är en kravspecifikation på ett inventeringssystem från lager- och skogsbranschen. Resultatet innehåller även tekniska lösningar som kan tillämpas och därmed uppfylla kraven. Parallellt med studien utvecklades en inventeringsklient tillsammans med Sweco som uppfyller både deras krav på klienten och de krav som studiens empiri erhållit. Resultatet av studien visar på ett system som klarar av att skapa, spara och återskapa dynamiska formulär. När formulär skapas kan dess kontrollelement ha restriktioner som klienten kan validera innan resultat sparas. Studiens resultat visar även på att en förstudie av ett system tenderar till att hitta indirekta/dolda krav.Den primära begränsningen i studien har varit tid. Hade mer tid kunnat ägnas åt studien skulle insamlingen av empiri ha varit mer omfattande. / Stocktaking is an exhausting process where employee’s collects data and is something that most companies do, although what they are collecting differs from business to business. Because it is a resource-waste to solve same problem multiple times, without reusing previous solutions, the student have choosen to study an overall solution for a stocktaking system that can be used by multiple businesses.The purpose of this report is to establish the end-user requirements on a stocktaking system and study which technical solutions can be used when developing a stocktaking system.The study used an abductive method with a case study on stocktaking. For the case study the empiricism was collected with interviews. A meta-analysis was implemented for the purpose to seek and analyse relevant literature.The result of the study was a requirement specification for an inventory system based on the forest and stock industry. The result also contains technical solutions that can be applied on the system in order to fulfill the requirements. In parallel with the study an attempt was made to develop a prototype of a stocktaking client togheter with Sweco that met both their requirements and requirements that was obtained from the interviews. The results show that the stocktaking client can manage to create, save and recreate dynamic forms. When forms are created their controls also have the ability to specify restrictions. The client can interpret these restrictions and validate the inputs before saving results. The results of the study also implies that a feasibility study of a system tend to find hidden requirements.The primary limitations of the study were time. If the study had a larger timescope more time could have been spent on collecting empirical data and gather end-user requirements.

Service composition integrates existing services to fulfill specific tasks using a set of standards and tools. Existing service composition techniques and tools are mainly designed for SOA professionals. The business processes used in the service composition systems are primarily designed by experienced business analysts who have extensive process knowledge. Process knowledge is the information about a process, including the tasks in a process, the control flow and data flow among tasks. It is challenging for end-users without sufficient service composition skills and process knowledge to find desired services then compose services to perform their daily activities, such as planning a trip. Context-aware techniques provide a promising way to help end-users find services using the context of end-users. However, existing context-aware techniques have limited support for dynamic adapting to new context types (e.g., location, time and activity) and context values (e.g., “New York City”). To shelter end-users from the complexity of service composition, we present our techniques that assist non-IT professional end-users in service composition by dynamically composing and recommending services to meet their requirements. To acquire the desired process knowledge for service composition, we propose an approach to automatically extract process knowledge from existing commercial applications on the Web. By analyzing the context of end-users, our techniques can dynamically adapt to new context types or values and provide personalized service recommendation for end-users. Instead of requiring end-users to specify detailed steps for service composition, the end-users only need to describe their goals using a few keywords. Our approach expands the meaning of an end-user's goal using process knowledge then derives a group of tasks to help the end-user fulfill the goal. The effectiveness of our proposed techniques is demonstrated through a set of case studies. / Thesis (Ph.D, Computing) -- Queen's University, 2011-09-30 11:43:39.151

The creation of interactive computer-controlled characters in interactive media is a challenging and multi-faceted task requiring the skills and effort of professionals from many fields. This work addresses authoring the interactive aspect of these characters’ behaviors – how characters act automatically in response to a dynamic user-controlled character. We present PaintBoard, a system that enables users to prototype and test discrete, real-time, interactive movements in a 2D grid environment simply by digitally painting a storyboard. We designed and developed a novel authoring technique for creating behaviors (painting storyboards) and a novel algorithm based on machine-learning, that analyzes a storyboard to create a behavior that works beyond situations provided in the input storyboard. We conducted two exploratory studies that grounded the prototype design, and present the results of a proof-of-concept workshop with game developers. Finally, we performed a comparison of machine learning algorithms’ performance on our storyboard data.

World Wide Web (WWW) has become the greatest repository of information that man has ever assembled and it is continuously growing. WWW transformed itself into a generative environment that fosters innovation through the advance of technologies and a shift in people's perception of the Weband how they use it. The new WWW or Future Internet is that of an Internet of Services and Internet of Things.Naturally, a series of questions arise from this context: how do you filter things to create more value than you currently get? how do you aggregate things in an intelligent and easy way instead of doing it in your head? The world cannot be described unambiguously, so how can you allow users to deal with the world in their own way, based on their understanding? Levine in his book "Cluetrain manifesto" was arguing that markets are conversations so how can users be involved in the conversation? how can users be empowered with easy consumption of the services, information, things that they found around?However design and deployment of such software capable of direct interaction and empowerment of the end-user is still an issue. We have on one side users that have ideas, but do not have technical background and lack programming skills to do the development by themselves. On the other side, we have large amounts of data, resources and services that could be aggregated both in terms of data, but most important in terms of behavior to innovate and create new things.Our goal in this thesis is to address this lack of tools that are capable of direct interaction andempowerment of end-users, in a unified manner. Thus our main contribution in this thesis is the development of a holistic approach for web based systems that are user-centric and that integrate data, services and behavior available on the Web 2.0.

With current state-of-the-art approaches, development of a single mobile manipulation capability can be a labor-intensive process that presents an impediment to the creation of general purpose household robots. At the same time, we expect that involving a larger community of non-roboticists can accelerate the creation of new novel behaviors. We introduce the use of a software authoring environment called ROS Commander (ROSCo) allowing end-users to create, refine, and reuse robot behaviors with complexity similar to those currently created by roboticists. Akin to Photoshop, which provides end-users with interfaces for advanced computer vision algorithms, our environment provides interfaces to mobile manipulation algorithmic building blocks that can be combined and configured to suit the demands of new tasks and their variations. As our system can be more demanding of users than alternatives such as using kinesthetic guidance or learning from demonstration, we performed a user study with 11 able-bodied participants and one person with quadriplegia to determine whether computer literate non-roboticists will be able to learn to use our tool. In our study, all participants were able to successfully construct functional behaviors after being trained. Furthermore, participants were able to produce behaviors that demonstrated a variety of creative manipulation strategies, showing the power of enabling end-users to author robot behaviors. Additionally, we introduce how using autonomous robot learning, where the robot captures its own training data, can complement human authoring of behaviors by freeing users from the repetitive task of capturing data for learning. By taking advantage of the robot's embodiment, our method creates classifiers that predict using visual appearances 3D locations on home mechanisms where user constructed behaviors will succeed. With active learning, we show that such classifiers can be learned using a small number of examples. We also show that this learning system works with behaviors constructed by non-roboticists in our user study. As far as we know, this is the first instance of perception learning with behaviors not hand-crafted by roboticists.

Organizations today are adopting business intelligence (BI) systems at a fast pace with the expectation that these systems will help them make better business decisions and improve their performance. But with the ever changing industry dynamics and variable business needs of end-users, the BI requirements of organizations are also getting unpredictable and increasingly harder to deliver by the service providers. Additionally, the advancement of web 2.0 & enterprise 2.0 technologies has opened up more possibilities for the development of user-centric innovative business applications. Enterprise BI Mashups are a specific breed of such technologies that have the potential to empower end-users with self-service capabilities and facilitate problem-solving in ad-hoc situational BI scenarios. This research project attempts to explore the current landscape of Business Intelligence (BI) Mashups and to identify gaps in technology with respect to user requirements and corporate objectives. Through an empirical investigation of BI mashups use cases, specific issues and challenges associated with the use of mashups in BI have been ascertained. Working in collaboration with IBM Cognos, we have formulated a taxonomy and utility framework for Enterprise BI Mashups. The formulated taxonomy provides a basic framework for understanding the domain of BI mashups and is aimed to aid application development initiatives for creating BI mashups toolkits. The utility framework draws upon real-world use cases for BI Mashups as well as pertinent software design patterns that can facilitate the development of BI mashup tools and services. These frameworks are expected to advance an understanding of business process requirements that can be satisfied through the use of Enterprise BI Mashups, and also aid in the development of mashup toolkits targeted at BI end-users.

A multiplatform e-learning system is an e-learning system that can deliver learning content to different accessing devices such as PCs, PDAs and mobile phones. The main objective of the research is to formulate a framework for multiplatform e-learning systems. This thesis focuses on the formulation, competency and constitution of the multiplatform e-learning systems framework and the implementation of a multiplatform e-learning system. In conjunction with the main objective, the research also addresses the factors that influence learner satisfaction during their engagement with a multiplatform e-learning system. In addition, the research investigates the relationships between these factors in influencing learner satisfaction. The research also intends to validate the assertion that multiplatform e-learning systems are better than non-adaptive e-learning systems. A comparative evaluation between a traditional e-learning system and a multiplatform e-learning system from end user (learner) perspective was conducted. The evaluation instrument is based on multiplatform e-learning system questionnaires (MELQ). A total of forty participants took part in the evaluation. Four participants took part in the initial pilot evaluation while thirty six participants took part in the final evaluation. Data analysis and statistical results indicate that there are potential gains in learner satisfaction score in multiplatform e-learning systems over traditional e-learning systems. The results also show that the gain is most significant in mobile devices than in desktop PCs. Statistical analysis reveals that all the factors that influence the learner satisfaction are significant and they have different levels of influence over learner satisfaction. These factors can be further organized into primary factors and secondary factors. These findings and the methodology of evaluation can play an important role for e-learning systems designer to improve the adaptation process and to enhance the level of learner satisfaction in multiplatform e-learning systems.