The sharing of needles among injection drug users (IDUs) is a common route of Human Immunodeficiency Virus and Hepatitis C Virus transmission. Through the increased utilization of social network analysis, researchers have been able to examine how the interpersonal relationships of IDUs affect injection risk behaviour. This study involves a secondary analysis of data from a cross-sectional study of 156 IDUs from Winnipeg, Manitoba titled “Social Network Analysis of Injection Drug Users”. Multiple logistic regression analysis was used to assess the individual and the social network characteristics associated with needle sharing among the IDUs. Generalized Estimating Equations analysis was used to determine the injecting dyad characteristics which influence needle sharing behaviour between the IDUs and their injection drug using network members. The results revealed five key thematic findings that were significantly associated with needle sharing: (1) types of drug use, (2) socio-demographic status, (3) injecting in semi-public locations, (4) intimacy, and (5) social influence. The findings from this study suggest that comprehensive prevention approaches that target individuals and their network relationships may be necessary for sustainable reductions in needle sharing among IDUs. / February 2006

The purpose of this study was to examine how youth useFacebook and how Facebook perceives construct their own identity. Anotherquestion was if the construction of the self-identity is different between thegenders.The study is based on empirical material that has been collected by usingquestionnaires. The questionnaire was handed out in three different classes in3rd grade in secondary school in southern Sweden. The analysis in this studywas based on two different kinds of theories, Erving Goffman’s social theory ofdramaturgical analysis and stigma and Yvonne Hirdman’s gender studies. The result showed that Facebook is a tool for youth to communicate and get intouch with other people. Youth uses Facebook to show selected parts ofthemselves, which create a ”hoped- for possible selves”. The result also showedthat girls are more aware of what they do on Facebook and the consequences of their activity,compared to boys. It seems that the construction of identity on Facebook variesbetween genders due to social structures.

Feature selection based on feature ranking has received much attention by researchers in the field of text classification. The major reasons are their scalability, ease of use, and fast computation. %, However, compared to the search-based feature selection methods such as wrappers and filters, they suffer from poor performance. This is linked to their major deficiencies, including: (i) feature ranking is problem-dependent; (ii) they ignore term dependencies, including redundancies and correlation; and (iii) they usually fail in unbalanced data. While using feature ranking methods for dimensionality reduction, we should be aware of these drawbacks, which arise from the function of feature ranking methods. In this thesis, a set of solutions is proposed to handle the drawbacks of feature ranking and boost their performance. First, an evaluation framework called feature meta-ranking is proposed to evaluate ranking measures. The framework is based on a newly proposed Differential Filter Level Performance (DFLP) measure. It was proved that, in ideal cases, the performance of text classifier is a monotonic, non-decreasing function of the number of features. Then we theoretically and empirically validate the effectiveness of DFLP as a meta-ranking measure to evaluate and compare feature ranking methods. The meta-ranking framework is also examined by a stopword extraction problem. We use the framework to select appropriate feature ranking measure for building domain-specific stoplists. The proposed framework is evaluated by SVM and Rocchio text classifiers on six benchmark data. The meta-ranking method suggests that in searching for a proper feature ranking measure, the backward feature ranking is as important as the forward one. Second, we show that the destructive effect of term redundancy gets worse as we decrease the feature ranking threshold. It implies that for aggressive feature selection, an effective redundancy reduction should be performed as well as feature ranking. An algorithm based on extracting term dependency links using an information theoretic inclusion index is proposed to detect and handle term dependencies. The dependency links are visualized by a tree structure called a term dependency tree. By grouping the nodes of the tree into two categories, including hub and link nodes, a heuristic algorithm is proposed to handle the term dependencies by merging or removing the link nodes. The proposed method of redundancy reduction is evaluated by SVM and Rocchio classifiers for four benchmark data sets. According to the results, redundancy reduction is more effective on weak classifiers since they are more sensitive to term redundancies. It also suggests that in those feature ranking methods which compact the information in a small number of features, aggressive feature selection is not recommended. Finally, to deal with class imbalance in feature level using ranking methods, a local feature ranking scheme called reverse discrimination approach is proposed. The proposed method is applied to a highly unbalanced social network discovery problem. In this case study, the problem of learning a social network is translated into a text classification problem using newly proposed actor and relationship modeling. Since social networks are usually sparse structures, the corresponding text classifiers become highly unbalanced. Experimental assessment of the reverse discrimination approach validates the effectiveness of the local feature ranking method to improve the classifier performance when dealing with unbalanced data. The application itself suggests a new approach to learn social structures from textual data.

Fueled by ubiquitous access to mobile phones, and a massive population of nearly 3 billion unbanked people around the globe, mobile commerce is evolving as a disruptive technology. Simultaneously, mobile payments are surfacing as a killer application within the mobile commerce context (Hu et al. 2008). Undeniably, the proliferation of wireless mobile technology provides much-needed access to vital information, and financial services for disenfranchised, unbanked populations. In addition, technological innovations offer first-time opportunities for suppliers of goods and services in a market context to gain competitive advantages while enhancing their economic viability. According to Portio Research, the volume of mobile payments rose significantly from $68.7 billion U.S. dollars in 2009, with predictions of $633.4 billion U.S. dollars by the end of 2014 (mobithinking.com 2012). Despite exponential growth in the number of mobile subscribers globally, and widespread adoption of mobile commerce, acceptance rates for mobile payment applications have lagged (Dahlberg et al. 2007, Ondrus et al 2009, Ondrus and Lyytinen 2011). Yet examinations of factors inhibiting the widespread acceptance of mobile payments are relatively sparse. Using Social Network theory, this research examines factors affecting engagement in mobile payments. The researcher posits that four primary elements influence mobile payment engagement: 1) the relationships between and amongst inter-organizational alliance members; 2) the prevailing regulatory environment; 3) the state of existing banking and technology infrastructures, and 4) an assessment of economic opportunity. The research outcomes from this exploratory examination led to the development of a comprehensive model for mobile payment engagement, and strongly suggest that ties between and amongst firms in inter-organizational alliances help ensure the success of mobile payment engagement. Support was found for the following suppositions: 1) similarities and relations (continuous ties) help establish a framework and understanding amongst alliance members as to each party’s goals and objectives; and 2) interactions and flows (discrete ties) between and amongst inter-organizational alliance members strengthen the overall ties between alliance members while solidifying a viable working relationship amongst the alliance members. This study employs a qualitative approach to obtain real world insight into the dynamism of the mobile payment arena. A model is then proposed to practically examine mobile payment engagement opportunities. From a theoretical perspective, the research contributes to the extant scholarly knowledgebase pertaining to engagement in mobile payments.

The prevalence of HIV/AIDS in the United States is still high despite advances in prevention and therapies. Among all races and ethnic groups, African Americans are the most severely affected and face a disproportionate burden. African Americans account for a higher proportion of HIV infections and deaths than other races and ethnicities. In addition, one of the fastest growing segments of AIDS cases is patients more than 50 years of age. This segment receives little attention concerning HIV infection and as the U.S. population continues to age, it is important to be aware of specific HIV-related risks faced by these older African Americans and to ensure that they get information and services to help protect them from infection. This study aims to understand and compare the social network characteristics, perceived risk of getting HIV infection and HIV risk behaviors between younger (18 to 49 years of age) and older (50 plus years of age) African Americans living in high HIV prevalence zip codes of Atlanta, Georgia. The study population included 897 African Americans. Controlling for socio-demographic variables, multivariate analyses revealed that older African Americans have significant higher proportion of injection drug use, are less likely to get tested for HIV and more likely to have a risky sex partner (i.e., exchange sex for money or drugs); however, older African Americans were less likely to engage in sexual risk behaviors. Groups did not differ in terms of their perceived risk for HIV and social network characteristics. More research is necessary to understand their HIV-related risk behaviors, both sexual and drug use, and the specific needs for primary prevention effort of HIV/AIDS transmission among older African Americans.

This dissertation research is focused on how we, as researchers, ‘influence’ others researchers. In particular, I am concerned with the notion of what constitutes the ‘influence’ of a scholar and how ‘influence’ is conferred upon scholars. This research is concerned with the construct called ‘scholarly influence’. Scholarly influence is of interest because a clear “theory of scholarly influence” does not yet exist. Rather a number of surrogate measures or concepts that are variable are used to evaluate the value of one’s academic work. ‘Scholarly influence’ is broken down into ‘ideational influence’ or the influence that one has through publication and the uptake of the ideas presented in the publication, and ‘social influence’ or the influence that one has through working with other researchers. Finally through the use of the definition of ‘scholarly influence’ this dissertation tries to commence a definition of ‘quality’ in scholarly work.

No description available.

Coworking is a new type of work model that has been developing rapidly during the 21th century. However, no academic research has been conducted on the subject yet. We have, therefore, decided to study the subject in order to evaluate if and how Coworking has an effect on creativity. The problem with this thesis is that creativity is a broad subject that includes several variables. This has led us to study theories within the field of: innovation, motivation, personal traits, and environment in respect to the Creative Process by Sawyer (2006). These theories have laid the foundation of our theoretical framework and are used to study our purpose and answer the research questions. This thesis is built upon data from individual case studies from semi-structured interviewees with coworkers from The Hub in Stockholm, Sweden. These interviews have then been transcribed and categorized by the Content Analysis in accordance with Hancock (1998).  The data has then been analyzed in-depth by using Eisenhardt’s Cross-Case Pattern Analysis (1989) in order to evaluate the relevance and reliability of the data. The results from our analysis are presented in unity with our method and theoretical framework, this part concludes with a reflection over our purpose and suggestions for future areas of research. The main finding from the results is that Coworking has a positive effect on creativity. But, this is mainly due to the mix of people participating in Coworking. The diverse group of coworkers creates at network of knowledge located in an open atmosphere that simplifies the creation of new ideas. We believe that this thesis has contributed to the academic society as it currently is the only academic paper within the area of Coworking.

Feature selection based on feature ranking has received much attention by researchers in the field of text classification. The major reasons are their scalability, ease of use, and fast computation. %, However, compared to the search-based feature selection methods such as wrappers and filters, they suffer from poor performance. This is linked to their major deficiencies, including: (i) feature ranking is problem-dependent; (ii) they ignore term dependencies, including redundancies and correlation; and (iii) they usually fail in unbalanced data. While using feature ranking methods for dimensionality reduction, we should be aware of these drawbacks, which arise from the function of feature ranking methods. In this thesis, a set of solutions is proposed to handle the drawbacks of feature ranking and boost their performance. First, an evaluation framework called feature meta-ranking is proposed to evaluate ranking measures. The framework is based on a newly proposed Differential Filter Level Performance (DFLP) measure. It was proved that, in ideal cases, the performance of text classifier is a monotonic, non-decreasing function of the number of features. Then we theoretically and empirically validate the effectiveness of DFLP as a meta-ranking measure to evaluate and compare feature ranking methods. The meta-ranking framework is also examined by a stopword extraction problem. We use the framework to select appropriate feature ranking measure for building domain-specific stoplists. The proposed framework is evaluated by SVM and Rocchio text classifiers on six benchmark data. The meta-ranking method suggests that in searching for a proper feature ranking measure, the backward feature ranking is as important as the forward one. Second, we show that the destructive effect of term redundancy gets worse as we decrease the feature ranking threshold. It implies that for aggressive feature selection, an effective redundancy reduction should be performed as well as feature ranking. An algorithm based on extracting term dependency links using an information theoretic inclusion index is proposed to detect and handle term dependencies. The dependency links are visualized by a tree structure called a term dependency tree. By grouping the nodes of the tree into two categories, including hub and link nodes, a heuristic algorithm is proposed to handle the term dependencies by merging or removing the link nodes. The proposed method of redundancy reduction is evaluated by SVM and Rocchio classifiers for four benchmark data sets. According to the results, redundancy reduction is more effective on weak classifiers since they are more sensitive to term redundancies. It also suggests that in those feature ranking methods which compact the information in a small number of features, aggressive feature selection is not recommended. Finally, to deal with class imbalance in feature level using ranking methods, a local feature ranking scheme called reverse discrimination approach is proposed. The proposed method is applied to a highly unbalanced social network discovery problem. In this case study, the problem of learning a social network is translated into a text classification problem using newly proposed actor and relationship modeling. Since social networks are usually sparse structures, the corresponding text classifiers become highly unbalanced. Experimental assessment of the reverse discrimination approach validates the effectiveness of the local feature ranking method to improve the classifier performance when dealing with unbalanced data. The application itself suggests a new approach to learn social structures from textual data.

New social networking and social web tools are becoming available and are easing the process of customizing online social environments. With these developments in technology, core design efforts are being extended beyond usability for individual users and beginning to include notions of sociability for the engagement of communities of users. This thesis is an investigation of these developments. It is guided by the principal research question: how do you design for social engagement in an online social environment intended to facilitate interaction in a community of users? To address this question, this thesis presents a domain-community model developed from the communities of practice concept and the Work Domain Analysis model used in Cognitive Work Analysis. The domain-community model provides a basis for the design a composition of web components for an online social environment that will addresses issues of social engagement and domain effectiveness. In a case study, the domain-community model was used as a basis for the redesign of a social networking portal used by an international development leadership community called UCP-SARnet. A social network analysis of core members of UCP-SARnet was conducted before and after the portal was redesigned. From the social network analysis, it was concluded that the structure of UCP-SARnet was positively affected by the redesign: core group members reported they knew one another significantly more after the redesign of the website than before the redesign. User experience measures of the UCP-SARnet portal, website usage data, and a tally of website communication activity also changed significantly with the redesign of the website. This provided more evidence that a design informed by Cognitive Work Analysis and communities of practice produced a measurable effect on the structure of the UCP-SARnet online community. As such, this model can provide a basis for designers of online communities to more systematically account for social phenomena in relation to collective efforts in a given work domain. Furthermore, it is expected the effectiveness of the model can be taken forward with future work by refining the domain-community model, developing techniques to translate the model into interface concepts, and building practices for community-based research and design.