abstract: Cyberbullying is a phenomenon which negatively affects individuals. Victims of the cyberbullying suffer from a range of mental issues, ranging from depression to low self-esteem. Due to the advent of the social media platforms, cyberbullying is becoming more and more prevalent. Traditional mechanisms to fight against cyberbullying include use of standards and guidelines, human moderators, use of blacklists based on profane words, and regular expressions to manually detect cyberbullying. However, these mechanisms fall short in social media and do not scale well. Users in social media use intentional evasive expressions like, obfuscation of abusive words, which necessitates the development of a sophisticated learning framework to automatically detect new cyberbullying behaviors. Cyberbullying detection in social media is a challenging task due to short, noisy and unstructured content and intentional obfuscation of the abusive words or phrases by social media users. Motivated by sociological and psychological findings on bullying behavior and its correlation with emotions, we propose to leverage the sentiment information to accurately detect cyberbullying behavior in social media by proposing an effective optimization framework. Experimental results on two real-world social media datasets show the superiority of the proposed framework. Further studies validate the effectiveness of leveraging sentiment information for cyberbullying detection. / Dissertation/Thesis / Masters Thesis Computer Science 2017

abstract: When preparing for and responding to disasters, humanitarian organizations must run effective and efficient supply chains to deliver the resources needed by the affected population. The management of humanitarian supply chains include coordinating the flows of goods, finances, and information. This dissertation examines how humanitarian organizations can improve the distribution of information, which is critical for the planning and coordination of the other two flows. Specifically, I study the diffusion of information on social media platforms since such platforms have emerged as useful communication tools for humanitarian organizations during times of crisis. In the first chapter, I identify several factors that affect how quickly information spreads on social media platforms. I utilized Twitter data from Hurricane Sandy, and the results indicate that the timing of information release and the influence of the content’s author determine information diffusion speed. The second chapter of this dissertation builds directly on the first study by also evaluating the rate at which social media content diffuses. A piece of content does not diffuse in isolation but, rather, coexists with other content on the same social media platform. After analyzing Twitter data from four distinct crises, the results indicate that other content’s diffusion often dampens a specific post’s diffusion speed. This is important for humanitarian organizations to recognize and carries implications for how they can coordinate with other organizations to avoid inhibiting the propagation of each other’s social media content. Finally, a user’s followers on social media platforms represent the user’s direct audience. The larger the user’s follower base, the more easily the same user can extensively broadcast information. Therefore, I study what drives the growth of humanitarian organizations’ follower bases during times of normalcy and emergency using Twitter data from one week before and one week after the 2016 Ecuador earthquake. / Dissertation/Thesis / Doctoral Dissertation Business Administration 2018

Background: Social medias have contributed to big changes in the way firms and customers interact with each other. Due to the great amount of people using social medias, firms have understood the importance of incorporating social medias in their marketing strategy. However, due to increasing competition, customer engagement on social medias has become important. To generate customer engagement, firms must create engaging social media content that is worth the customer’s time and effort. The micro-sized firms are under researched and therefore the authors have chosen micro-sized firms as a study subject and hence the sender of the content. Purpose: The purpose of this thesis is to describe the key components of creating customer engagement on social medias for micro-sized firms. Methodology: This research took a qualitative approach in a deductive nature with a descriptive purpose and a cross-sectional research design. The need of primary data was collected by conducting seven semi-structured interviews with social media managers- and consultants. Participants were purposely sampled and the empirical findings were coded, concepts were identified and further categorised while trustworthiness was set to be a criterion for the quality of the research. Conclusion: The research provided seven key components of engaging social media content for micro-sized firms and further recommendations on how to implement them in the micro-sized firms social media management. The components are described together with an illustration. Keywords: Micro-sized firms, social media managers, social media, customer engagement, content appeals, social media targeting, social media timing and frequency, qualitative research, semi-structured interviews.

A profusion of research has been dedicated to the effects of new media on body image. In an attempt to explain the inflation of eating disorders, several researchers have turned to the cultivation theory, postulating that increased interaction with these medias will lead to internalizations of the messages they disseminate. The over presence of extremely thin models and actresses can create a new reality for media users, who begin to equate thinness with beauty, power, femininity and happiness. While an abundance of research has delved into the impact of this thin ideal through television and magazines, the Internet as a medium is relatively new territory. Untrodden further is the field of social media, and particularly Instagram. The aim of this study was to determine the potential of cultivation theory via Instagram and its users.

The aim of this research was creating metrics for measuring social connectedness in social media. This thesis made use of social capital theory in order to inform the construction of original metrics. The methodology used in this thesis involved conducting a literature review into the use of social capital theory in social media, proposing new metrics, implementation in software, validation, evaluation against other measures and finally demonstrating the utility of the new metrics. A preliminary case study verified the suitability of using Facebook as a context for developing the metrics. The main practical work outlined in this thesis aimed to validate Social Capital in Social Media (SCiSM) metrics against the Internet Social Capital Scale (ISCS) (Williams 2006). The SCiSM metrics were developed to relate to bonding social capital, bridging social capital and total social capital (Putnam 2000). The methodology used to validate the SCiSM metrics was Meneely (2012) and involved using two independent data sets to validate the SCiSM metrics using both correlations and linear regression. Statistical analysis found a strong positive correlation between ISCS and SCiSM whilst regression analysis demonstrated that the relationship between SCiSM and ISCS was concerned with ranking rather than an absolute number. SCiSM was evaluated against other social capital metrics used in the literature such as degree centrality. It was found that SCiSM had a higher number of significant correlations with the ISCS than other measures. The SCiSM metrics were then used to analyse the two independent data sets in order to demonstrate their utility. The first data set, taken from a Facebook group, was analysed using a paired t-test. It was found that bonding social capital increased over a twelve week period but that bridging social capital did not. The second data set, which was taken from Facebook status updates, was analysed using correlations. The result was that there was a positive correlation between number of Facebook friends and bonding social capital. However it was also found that there was a negative correlation between number of Facebook friends and bridging social capital. This suggests that there is a dilution effect in the usefulness of large friend networks for bridging social capital. In conclusion the problem that this research has addressed is providing a means to improve understanding of social capital in social media.

University of Macau / Faculty of Science and Technology / Department of Computer and Information Science

The purpose of this project is to examine the visual rhetoric of one online community. Drawing heavily from the work of Laurie Gries (2015), I track the evolution of an image as it circulates through a forum of photo manipulators in the group “PhotoShopBattles.” While Gries’ work traced the evolution of the iconic Obama Hope poster and its iterations in various media, this project restricts its observations to the images posted to one webpage, focusing on one evolutionary chain. By narrowing the focus to one internet forum page that evolved over the course of one week, we can observe linear evolution that occurs quite rapidly. While the study of Obama Hope covered years, the works in this study were constructed in a matter of days. Additionally, the site records the step-by-step progress of the reformed work. The Obama Hope work offers guidance to the work in this project. Using this method, an image’s evolution is broken down into steps and principles noted in the life of the image or multiple-image. By recreating a slightly modified version of Gries’ method, this work seeks to decode the meanings and outcomes that are created and changed around one set of evolving images. New materialism offers a way to explore the visual rhetorical moves of members of an online community and discuss the outcomes associated with those moves. With a deeper understanding of the environment surrounding image creation and the outcomes derived from an image, we can better understand the way image is used rhetorically online.

Given the importance of a business being able to create a positive brand image in the minds of customers and influencing customer purchasing behaviour, there is a lack of attention given to the research of the role that social media tools can play in the success of retailers. Numerous research studies have focussed on determining the advantages of and impact that social media tools can have on the success of businesses, however, the impact of social media tools on the creation of a positive brand image, and influencing customer purchasing behaviour is less evident. Thus, the primary objective of this study is to investigate the influence of social media tools on the brand image of retailers and how it affects customer purchasing behaviour. A full literature review was conducted on the retail environment and its importance, as well as on social media as a marketing communication tool. Thereafter, independent variables (Official websites, Facebook and Twitter) were identified as possibly influencing the intervening variable (Brand image) and dependent variable (Customer purchasing behaviour). Three independent variables were selected and used to construct a hypothetical model and research hypothesis. The purpose of these independent variables was to determine whether the use of social media tools could lead to a positive brand image and ultimately improve customer purchasing behaviour. An empirical investigation was undertaken to be able to establish the influence of social media tools on the brand image and customer purchasing behaviour of retail customers. The measuring instrument used for this study, a questionnaire, was constructed from secondary literature sources. The convenience sampling technique was used to identify respondents. The usable questionnaires that were gathered from respondents were examined through statistical analyses. Exploratory factor analysis (EFA) and the calculation of Cronbach’s alpha coefficients were used to confirm the validity and reliability of the measuring instrument. Descriptive statistics, as part of the data analyses, were calculated to summarise and allow interpretation of the sample data. Furthermore, Pearson’s Product Moment Correlations were calculated to determine correlations among variables used in this study. Multiple regression analysis was the primary statistical procedure used for testing the significance of relationships hypothesised between the independent, intervening and dependent variables of this study. The empirical investigation was concluded by a the calculation of a t-test and Analysis of Variance (ANOVA), for the purpose of determining the influence of selected demographic variables on respondents’ perceptions with regard to the independent, intervening and dependent variables. In addition, post-hoc Tukey tests were conducted to determine significant differences among individual mean scores, and the degree of practical significance was determined by calculated Cohen’s d values. Based on the results of the empirical investigation, significant relationships were found among the independent variables (Official websites, Facebook and Twitter), the intervening variable (Brand image) and the dependent variable (Customer purchasing behaviour). The empirical investigation revealed that the age of respondents has a significant impact on how their image of a brand and purchasing behaviour are influenced by retailers’ use of social media tools. This study has made a contribution to the shortage of literature on the impact of social media tools on retailers. The hypothetical model developed for this study assisted in understanding the impact that retailers’ use of social media tools have on the brand image of retailers and customers’ purchasing behaviour. As a result, this study provides recommendations and suggestions for retailers to ensure a positive brand image in the minds of customers, and to ultimately use social media tools to positively influence the purchasing behaviour of customers.

University sustainability is a field of research that has been gaining increased interest in recent years. The reduction of environmental impact has become a strategic objective of universities globally. Universities have been prompted to take necessary action to ensure that their environmental impact is at a minimum. The environmental component of sustainability deals with the current conservation of the earth’s natural resources so that future generations can also have access to them. Human beings, due to their increasing needs, are accountable for the exploitation of natural resources. They are regarded as the main contributors to imbalances in the natural systems. Environmental concerns such as global warming, deforestations, disposal of wastes, and ozone reduction are the outcomes of the damage caused by humans on the environment. The aim of environmental education is to acquire remediation of the environment by making individuals aware of the environment and by educating them about how to live a more sustainable lifestyle. Environmental awareness is perceived as knowledge of the factors that affect the environment and having sensitivity towards the environment. Higher Education Institutions (HEIs) bear the responsibility of educating individuals about environmental issues since they provide education to future leaders in society who may have an influence on future conditions in the environment. Social media are capable of delivering information to a large spectrum of audiences at a low cost. The Pew Internet American Life Project reported that the number of adults who utilise social media has increased by 57 percent from 2005 to 2011. Several environmental activist organisations utilise social media to carry out environmental awareness campaigns. In this study two environmental awareness campaigns which were powered by social media were conducted to improve environmental awareness of individuals in a higher education environment. A Social media Model for ENvironmental Awareness (SMENA) was developed to facilitate the environmental awareness campaigns. The SMENA includes a website, social media as well as theoretical guidelines for creating environmental awareness campaigns, and for using social media for environmental awareness campaigns. A case study at the Nelson Mandela Metropolitan University (NMMU) was used to empirically evaluate SMENA. Students at the Department of Computer Sciences of NMMU were exposed to information about environmental issues through social media with the intention of improving their environmental knowledge and awareness. The SMENA website usability was rated positively and students enjoyed the blogs and information distributed by means of social media. The results of the study intervention were positive and showed that social media can be used to improve the environmental knowledge of students. This study provides a valuable contribution to both the field of environmental education and social media usage and acceptance. The guidelines and requirements for using social media to improve environmental awareness provided in this study can be used to assist educators and university management with addressing the problems of reducing environmental impact.

This thesis examines the use of social media by museums aiming to establish collaborative relationships with the public. Social media platforms have been widely espoused as transformative in allowing diverse, new or previously excluded audiences to enter into egalitarian, participatory relationships with museums. This thesis deconstructs the concepts of participation and collaboration and identifies the various factors that constrain the extent to which social media enables participatory relationships between previously unequal actors. These factors include the historical disciplinary aims and cultural authority of museums, persistent social inequalities, and the motivations of social media followers. It elucidates crucial questions such as, are various publics enabled to participate on an equal level with each other and with museums? Who benefits from collaborative projects in general and which parties benefit from the use of social media in particular? What are the factors that limit the establishment of collaborative practice? And, conversely, what are the factors that define truly collaborative practice? This research examines museums' use of and discourses surrounding social media as well as social media followers' motivations for engaging with museums online. A large body of quantitative and qualitative data gained through in-depth web-based surveys is analysed, primarily using critical discourse analysis, and informed by other critical orientations including media archaeology and the sociology of expertise. The analysis indicates that museums consider social media to be a transformative, democratising technology. However, museums' acceptance of technologically determinist arguments significantly inhibits positive societal change and the extent to which collaborative relationships can be established with various publics. This research contributes significantly to the existing archaeological and museum studies literature by providing a theoretically and empirically informed critical analysis of the prevailing positive discourses surrounding social media and participation. It has important practical implications for museums in arguing that targeted, critically informed and ethically aware projects are necessary to achieve situations resembling 'collaboration'. It provides a significant body of data that will inform the formulation and continuation of collaborative projects in museums. Furthermore, it informs broader archaeological debates on involving various publics in archaeological practice. This thesis also demonstrates the importance and effectiveness of critical discourse analysis and related critical approaches for analysing large bodies of qualitative data.