Submitted by Johnny Rodrigues (johnnyrodrigues@ufcg.edu.br) on 2018-05-02T15:02:59Z No. of bitstreams: 1 LENIN DA NÓBREGA MEDEIROS - DISSERTAÇÃO PPGCC 2015..pdf: 824404 bytes, checksum: fe11b50d651b97b09c377e061155f477 (MD5) / Made available in DSpace on 2018-05-02T15:02:59Z (GMT). No. of bitstreams: 1 LENIN DA NÓBREGA MEDEIROS - DISSERTAÇÃO PPGCC 2015..pdf: 824404 bytes, checksum: fe11b50d651b97b09c377e061155f477 (MD5) Previous issue date: 2015-07 / Sites Q&A são aqueles acessados por usuários com o intuito de: encontrar uma resposta para alguma pergunta ou ajudar alguém a conseguir isso. Muitos pesquisadores têm concentrado esforços visando descobrir maneiras de atrair mais conteúdo para sites desse tipo e, também, entender comportamentos e motivações dos respectivos usuários. Investigou-se, neste trabalho, a utilização da prática do compartilhamento de perguntas oriundas de sites Q&A em redes sociais como estratégia para fomentar a contribuição nesses sites. Nosso intuito era: entender como os usuários se comportam diante dessa prática, investigar como promovê-la e, por fim, tentar encontrar indícios acerca da utilidade dessa prática como meio de fomentar a contribuição em sites Q&A. Foram feitas duas investigações qualitativas nas quais usuários de sites Q&A e de redes sociais foram entrevistados e relataram como se comportam diante do compartilhamento de perguntas. Posteriormente, foi realizado um estudo de caso utilizando perguntas e usuários do site Q&A Stack Over flow em Português para testar diferentes formas de incentivar o compartilhamento dessas perguntas em redes sociais, bem como para investigar mais a fundo se essa prática ajuda em algo na obtenção de respostas para essas perguntas. Descobriu-se que,por causa dos custos sociais envolvidos, os usuários tendem a não querer compartilhar, nas suas redes sociais, perguntas oriundas de sites Q&A realizadas por terceiros e que estejam sem resposta. Por este motivo concluímos que é preciso repensar a forma com a qual se tenta promover o compartilhamento de perguntas de sites Q&A em redes sociais. Além disso, também foi possível elaborar a hipótese de que o compartilhamento de perguntas em redes sociais pode não ser uma boa escolha para quem pretende fomentar a contribuição em sites Q&A, pois isso talvez não sirva para aumentar as chances de uma pergunta ser respondida. / Q&A sites are those accessed by users who want to find answers to their questions or answer someone else’s questions. Programmers use these kind of website very much in order to exchange questions and answers regarding programming. This behavior was observed in a website called Stack Over flow emPortuguês. Many researchers have been focused on finding out ways to attract more content to those websites and also on the aspects of the behaviors and motivations of their users. In this work our focus was the the practice of sharing Q&A sites questions in social networks as a strategy to increase the content of these websites. Our goals were: extract some lessons about how the users ’behavior look like when they face the functionality of sharing questions in Q&A sites, investigate how to promote this behavior and find out statistical evidence with respect to the advantages of sharing questions in Q&A sites and if this practice helps to increase the number of answers. We conducted two qualitative investigations in which users from Q&A sites and social networks were interviewed and they reported how is their general behavior when they come across with the option of sharing questions in those websites. Then, we performed a quantitative experiment to test different ways of promoting the act of sharing questions from Q&A sites in social networks ,as well as to investigate more deeply the effect of this practice in these websites contents. We concluded that users are very discerning when deciding whether or not to share content in their social networks due to the related social costs. That’s why we believe that it’s necessary to change the current way of promoting the sharing of Q&A sites questions in social networks. Besides, we noticed that sharing of questions may not be a good strategy to increase the contribution in these kind of websites.

The ductility of Q&P 1180 steel was studied with regard to retained austenite transformation under different strain paths. Specimens were tested in uniaxial tension in a standard test frame as well as in situ in the scanning electron microscope (SEM). Then digital image correlation (DIC) was used to compute the effective strain at the level of the individual phases in the microstructure. Stretching experiments were also performed using limiting dome height (LDH) tooling, where specimens were strained in both biaxial and plane strain tension. The experiments were done incrementally, for each strain path, and the retained austenite at each level of strain was measured using electron backscatter diffraction (EBSD). Retained austenite levels in the uniaxial tension case dropped from an initial measured level of about 8% to about 2% during an initial strain increment of 0.02, but then stabilized as the specimen was strained to 0.1. In the plane strain and biaxial tension cases retained austenite also dropped significantly during an initial strain increment of about 0.04, but then continued to decrease as the specimens were strained to failure. Biaxial tension, in particular, was the most effective strain path for transforming retained austenite to martensite, resulting in a final volume fraction of 0.3% at an effective strain of 0.3. Retained austenite in the plane-strain tension case dropped at a faster rate than in the biaxial tension case, but finished at about 1% at a strain of 0.1. The greatest limit strains were seen in the biaxial tension case, which may be partly explained by the more effective conversion of retained austenite than was seen in the uniaxial tension case.

Many individuals with autism manifest problems in transitioning to adulthood due to social impairments, communication difficulties, and rigid behaviors. One of those challenges is developing a robust and sufficiently large network of people who can provide advice about a variety of everyday situations. In this dissertation, I investigate ways of supporting adolescents with high functioning autism in navigating their everyday life through specializing social networking services (SNSs). A series of studies were conducted to identify opportunities for the design and use of a specialized SNS to foster the independence. The results demonstrates SNS can support some adolescents and adults with autism in increasing the independence by 1) facilitating the asking of a variety of online networks beyond a primary caregiver; and 2) allowing acquisition of rapid, direct, and informational advice with crowdsourcing. Drawing on several formative studies and investigations, I synthesized design guidelines for inquir.us, a specialized hybrid social question-and-answer (Q&A) platform with features for scaffolding question creation and crowdsourcing answers. Through the initial evaluation of inquir.us, I examined the Q&A behavior of individuals with autism on this platform and identified both opportunities and barriers to adoption in the context of supporting transition skills for the independence. The contributions of this thesis are: (1) a rich description of challenges and opportunities related to attaining independent life using SNSs; (2) empirical studies of individuals with autism’s online Q&A behavior; (3) design implications for designing a specialized SNS facilitating the Q&A interactions; and (4) the design and exploratory study of a social Q&A platform in the real world.

abstract: Online discussion forums have become an integral part of education and are large repositories of valuable information. They facilitate exploratory learning by allowing users to review and respond to the work of others and approach learning in diverse ways. This research investigates the different comment semantic features and the effect they have on the quality of a post in a large-scale discussion forum. We survey the relevant literature and employ the key content quality identification features. We then construct comment semantics features and build several regression models to explore the value of comment semantics dynamics. The results reconfirm the usefulness of several essential quality predictors, including time, reputation, length, and editorship. We also found that comment semantics are valuable to shape the answer quality. Specifically, the diversity of comments significantly contributes to the answer quality. In addition, when searching for good quality answers, it is important to look for global semantics dynamics (diversity), rather than observe local differences (disputable content). Finally, the presence of comments shepherd the community to revise the posts by attracting attentions to the posts and eventually facilitate the editing process. / Dissertation/Thesis / Masters Thesis Computer Science 2016

Submitted by Lucienne Costa (lucienneferreira@ufcg.edu.br) on 2018-04-30T19:24:01Z No. of bitstreams: 1 NIGINI ABILIO OLIVEIRA – TESE (PPGCC) 2017.pdf: 1277477 bytes, checksum: a8ef0f0459a6aacf859888c83f4ffde5 (MD5) / Made available in DSpace on 2018-04-30T19:24:01Z (GMT). No. of bitstreams: 1 NIGINI ABILIO OLIVEIRA – TESE (PPGCC) 2017.pdf: 1277477 bytes, checksum: a8ef0f0459a6aacf859888c83f4ffde5 (MD5) Previous issue date: 2017-05-04 / Os sites de Perguntas e Respostas (Q&A) têm o objetivo de solucionar os problemas de seus participantes fornecendo ferramentas para que eles criem colaborativamente respostas para perguntas feitas, construindo assim um repositório de conhecimento pesquisável. Mais do que um repositório de informações, esses ambientes são comunidades de pessoas que interagem em torno do conhecimento criado. No entanto, como os colaboradores estão espalhados por todo o mundo, estes tendem a não compartilhar o mesmo contexto socioeconômico e cultural. Essas diferenças podem criar barreiras ou oportunidades de colaboração, e a falta de observação dessas diferenças pode resultar em comunidades menos diversificadas e produtivas. Esta pesquisa baseia-se na premissa de que ambientes online abertamente disponíveis podem ser projetados para apoiar igualmente o engajamento de comunidades culturalmente diversas e visa melhorar o conhecimento sobre como fazê-lo no caso de sites de Q&A. Embora o sucesso dos ambientes de perguntas e respostas online dependa da participação dos usuários, a literatura mostra que o número de contribuições varia entre os países, além de estarem associados a valores culturais regionais. Nós seguimos estes resultados para examinar: (1) se tais diferenças se mantêm em dois sites populares não explorados previamente; (2) Se as diferenças na participação dos grupos nacionais também ocorrem porque alguns países têm uma porcentagem maior de usuários que estão dispostos a contribuir; e (3) Quais valores e perspectivas culturais sobre a colaboração podem orientar a concepção de sites de Q&A e outros ambientes interculturais de produção de conhecimento colaborativo mais inclusivos. Para responder às nossas perguntas de pesquisa usamos uma abordagem de métodos mistos, começando por desenvolver uma exploração quantitativa da participação dos grupos nacionais e das explicações significativas para possíveis diferenças. Em seguida, mostramos como essas diferenças podem ser compreendidas através de um estudo qualitativo: uma comparação baseada nos valores humanos segundo a perspectiva dos projetistas do site e das preferências dos participantes de três países. Nossa análise quantitativa confirmou resultados anteriores sobre a relação entre cultura nacional – mais especificamente o construto “Individualismo versus Coletivismo” – e diferenças em participação online. / Question and Answer (Q&A) sites have the goal of solving participant’s problems by provid- ing tools for them to collaboratively create answers to posed questions and build a repository of searchable knowledge. More than a repository for information though, these environments are communities of people that interact around the created knowledge. However, because collaborators are spread around the world they tend not to share the same socioeconomic context nor the cultural background. Such differences can either create barriers or opportunities for collaboration, and a lack of observation of these differences might result in less diverse and productive communities. This research is grounded in the perspective that openly-available online environments can be designed to equally support the engagement of culturally diverse communities, and aims to improve the knowledge on how to do so in the case of Q&A sites. While the success of online Question & Answer environments relies on user participation, previous work has shown that the number of contributions varies between countries and that they are also associated with regional cultural values. We follow this lead to examine: (1) If such differences hold for two not previously explored popular Q&A sites; (2) Whether differences in national groups participation also happen because certain countries have a higher percentage of users who are willing to contribute; and (3) What local cultural values and perspectives on collaboration can guide the design of more inclusive Q&A sites and other knowledge-based intercultural peer-production activities. To answer our research questions we use a mixed-methods approach, starting by developing a quantitative exploration of national groups’ participation and the significant explanations for its differences. We then show how these differences can be further understood through a qualitative study, a human values based comparison of the site designers’ perspectives and the participants’ preferences from three national groups. Our quantitative analysis confirm previous results regarding the relation between national culture – more specifically the “Individualism versus Collectivism” construct – and differences in online participation. For instance, we find that the countries with a smaller percentage of participants who engage in answering or commenting to posts are more likely to be from countries that have been shown to be more collectivists (mostly the ones outside North America and Western Europe) or have lower English proficiency indexes. We complement this result with an analysis of interviews with Question & Answer site users from three countries: China, India and United States. Our results show that Indian and Chinese participants seem to search for more social interactions than Americans – an affordance that is in general not supported by the studied sites. This research surfaces the human values inherent to online Question & Answer communities and the existent tensions between different stakeholders collaborating to create high quality knowledge artifacts. Based on that we discuss how specific design decisions on these sites, such as the competitive reward mechanism used to encourage contributions could be changed to encourage currently passive people to contribute. By clarifying the nuanced information underlying the relation between cultural theories and online engagement we hope to contribute to the design of more culturally-aware online environments.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第21398号 / 情博第684号 / 新制||情||118(附属図書館) / 京都大学大学院情報学研究科社会情報学専攻 / (主査)准教授 松原 繁夫, 教授 吉川 正俊, 教授 西田 豊明 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

Creating websites today is not easy as it seems, therefore it is important to know the essential principle when developing new product for users. Knowing the essential principles is not enough, you will also need to know about what users think and feel is important on a website such as Stack Overflow. It is a question-and-answer website where it offers users the ability to post questions and receive answers from other users that uses the same website. This study will identify what is important for young adult developers on the website Stack Overflow and what they think, and feel is required in terms of usability on the website. In this study there will be a theoretical framework and it includes different aspects of usability: learnability, comprehensibility, navigation, flexibility, and discoverability. This theoretical framework has been created to be able to identify what is important for young adult developers on the website Stack Overflow. By using this theoretical framework, we can setup the data collection method which is interview and think-aloud-test. As for analysis of the data, content analysis has been used to analyze the results and the study reached its conclusion. The conclusion is that on a website like Stack Overflow, navigation, flexibility, and comprehensibility are the important aspects for young adult developers.

TBF 1180 steel was studied under various conditions focusing on the correlation of ductility and amount of retained austenite. Samples were prepared from sheet stock and then strained using limiting dome height tooling (LDH), a standard uniaxial test frame, and a tensile stage for use in an electron microscope. The steel was observed in plane, biaxial, and uniaxial strain to determine its effect on retained austenite transformation and ultimately, ductility. Retained austenite was observed using a scanning electron microscope (SEM) equipped with an electron backscatter detector (EBSD) to distinguish the different phases present. Initial austenite levels were around 5% by volume and was quickly reduced as the sample was strained. The biaxial samples were the slowest to transform, having about 2.5% austenite at .05 effective strain, which allowed the specimen to reach an effective strain of .3 with 1.1% austenite remaining. In contrast, the plane strain samples had the fastest rate of transformation having only 1.2% austenite at .05 effective strain and .7% austenite at a final effective strain of .18. Both forms of uniaxial, (in-situ and ex-situ), were near identical, as expected, and exhibited an austenite transformation curve between that of the plane and biaxial curves. The uniaxial austenite level at .05 strain was 2.1% and was able to reach about .15 strain with a final austenite percentage around 1%. It was concluded that the biaxial strain path had the greatest ductility due in part to its slower austenite transformation rate while plane and uniaxial strain paths were not as ductile with their faster austenite transformation rates.

This Bachelor thesis is about digital tools for event organisers. The platforms offered by companies serve as digital backchannels that among other things facilitate interactions between audience and stage, for example during live Q&A sessions. This work focuses on the tasks and challenges of moderators on site and aims to develop a system that will optimise their workflows. Therefore, a tool is developed that allows moderators to conduct a Q&A session independently without having to rely on the assistance of an event organiser operating a desktop based platform simultaneously. The analysis addresses the moderator's role and tasks. It turns out that the functions of the platform relevant to the moderator can be combined in a mobile interface that is linked to it. Using a semi- structured expert interview, the feedback from clients of the start-up "Happenn" is gathered, which use the already existing platform of the company. Their platform "Happenn Live" serves as case study for this thesis. On the basis of the findings of the methods adapted to Lean UX, a solution approach in the form of a redesign of the platform and a corresponding mobile application for moderators is conceptualised. The structure and visual design of the platform follows the general design standards for desktop- and mobile-based interfaces. In this way, an improved usability can be achieved by making the platform more user-friendly for the user.

abstract: This dissertation seeks to understand and study the process of attention harvesting and knowledge production on typical online Q&A communities. Goals of this study include quantifying the attention harvesting and online knowledge, damping the effect of competition for attention on knowledge production, and examining the diversity of user behaviors on question answering. Project 1 starts with a simplistic discrete time model on a scale-free network and provides the method to measure the attention harvested. Further, project 1 highlights the effect of distractions on harvesting productive attention and in the end concludes which factors are influential and sensitive to the attention harvesting. The main finding is the critical condition to optimize the attention harvesting on the network by reducing network connection. Project 2 extends the scope of the study to quantify the value and quality of knowledge, focusing on the question answering dynamics. This part of research models how attention was distributed under typical answering strategies on a virtual online Q&A community. The final result provides an approach to measure the efficiency of attention transferred into value production and observes the contribution of different scenarios under various computed metrics. Project 3 is an advanced study on the foundation of the virtual question answering community from project 2. With highlights of different user behavioral preferences, algorithm stochastically simulates individual decisions and behavior. Results from sensitivity analysis on different mixtures of user groups gives insight of nonlinear dynamics for the objectives of success. Simulation finding shows reputation rewarding mechanism on Stack Overflow shapes the crowd mixture of behavior to be successful. In addition, project proposed an attention allocation scenario of question answering to improve the success metrics when coupling with a particular selection strategy. / Dissertation/Thesis / Doctoral Dissertation Applied Mathematics for the Life and Social Sciences 2019