Adopting Learning Analytics to Inform Postgraduate Curriculum Design: Recommendations and Research Agenda

2023 | Information Systems Frontiers | Citations: 0

Authors: Dennehy, Denis; Conboy, Kieran; Babu, Jaganath

Abstract: Understanding student sentiment plays a vital role in understanding the changes ... Expand

Abstract: Understanding student sentiment plays a vital role in understanding the changes that could or should be made in curriculum design at university. Learning Analytics (LA) has shown potential for improving student learning experiences and supporting teacher inquiry. Yet, there is limited research that reports on the adoption and actual use of LA to support teacher inquiry. This four-year longitudinal study captures sentiment of postgraduate students at a university in Ireland, by integrating LA with the steps of teacher inquiry. This study makes three important contributions to teaching and learning literature. First, it reports on the use of LA to support teacher inquiry over four one-year cycles of a Master of Science in Business Analytics programme between 2016 and 2020. Second, it provides evidence-based recommendations on how to optimise LA to support teacher inquiry, with specific attention as to how these can improve the assimilation of LA into the curriculum design and delivery. Third, the paper concludes with a research agenda to help improve the adoption and integration of LA in the future. Collapse

Topics: innovation management Tableau software analytical information system interactive data visualization Python

Methods: qualitative interview Cross-Industry Standard Process for Data Mining sentiment analysis case study design framework

Theories: assimilation theory of learning

Toward a campus crisis management system amid the pandemic and beyond

2023 | Information & Management | Citations: 0

Authors: Li, Yaojie; Zhou, Yi; Ge, Linqiang; Chen, Rui; Xiong, Jie

Abstract: This study develops artifact and design principles for a campus-wide crisis mana ... Expand

Abstract: This study develops artifact and design principles for a campus-wide crisis management system amid the COVID19 pandemic and beyond. Drawing upon crisis management literature and the uncertainty reduction theory, we develop a preliminary crisis management system with three fundamental components: dashboard, portal, and forum, geared toward aggregating environmental information, expert knowledge, and stakeholder insights. This immediate, integrated, and interactive campus crisis management system aims to promote organizational learning through knowledge acquisition, dissemination, and utilization while reducing stakeholders’ uncertainty about the crisis. Through comprehensive evaluation, we further synthesize and summarize new design principles for prospective artifacts, including constructing a distributed crisis management network, capitalizing on data mining and knowledge discovery algorithms, and developing the adaptability of the crisis management system in various circumstances. Collapse

Topics: emergency management pandemic organization learning digital dashboard usability

Methods: design principle design science design artifact machine learning kernel theory

Theories: uncertainty reduction theory organizational learning theory theory of abduction

Visual Analytics: Transferring, Translating and Transforming Knowledge from Analytics Experts to Non-technical Domain Experts in Multidisciplinary Teams

2023 | Information Systems Frontiers | Citations: 0

Authors: Marjanovic, Olivera; Patmore, Greg; Balnave, Nikola

Abstract: Today’s complex problems call for multidisciplinary analytics teams comprising o ... Expand

Abstract: Today’s complex problems call for multidisciplinary analytics teams comprising of both analytics and non-technical domain (i.e. subject matter) experts. Recognizing the difference between data visualisaion (DV) (i.e. static visual outputs) and visual analytics (VA) (i.e. a process of interactive visual data exploration, guided by user’s domain and contextual knowledge), this paper focuses on VA for non-technical domain experts. By seeking to understand knowledge sharing from VA experts to non-technical users of VA in a multidisciplinary team, we aim to explore how these domain experts learn to use VA as a thinking tool, guided by their knowing-in-practice. The research described in this paper was conducted in the context of a long-term industry-wide research project called the ‘Visual Historical Atlas of the Australian Co-operatives’, led by a multidisciplinary VA team who faced the challenge tackled by this research. Using Action Design Research (ADR) and the combined theoretical lens of boundary objects and secondary design, the paper theorises a three-phase method for knowledge transfer, translation and transformation from VA experts to domain experts using different types of VA-related boundary objects. Together with the proposed set of design principles, the three-phase model advances the well-established stream of research on organizational use of analytics, extending it to the emerging area of visual analytics for non-technical decision makers. Collapse

Topics: knowledge sharing data visualization implicit knowledge knowledge management participatory design

Methods: kernel theory design science action design research exploratory data analysis theory development

Theories: boundary objects theory

Monitoring and Understanding Enterprise Collaboration Platform Outcomes and Benefits Change

2020 | HICSS | Citations: 2

Authors: Nitschke, Clara S; Williams, Susan P

Abstract: Enterprise collaboration platforms integrating traditional collaboration tools a ... Expand

Abstract: Enterprise collaboration platforms integrating traditional collaboration tools and enterprise social software are shaped and designed through use. To date, existing research has not studied in any depth how their outcomes and benefits change over time. In this paper, we develop the MoBeC framework for capturing and monitoring how outcomes and benefits of enterprise collaboration platforms are changing over time. The framework is applied in an empirical setting adopting a longitudinal case study design. The study findings contribute to the deeper understanding of the dynamic and evolving nature of such platforms. Collapse

Topics: social media IS impact IT supported collaboration IT investment IS success

Methods: case study longitudinal research qualitative interview longitudinal case study literature study

Theories: theory of affordance

Exploring adverse drug reactions of diabetes medicine using social media analytics and interactive visualizations

2019 | International Journal of Information Management | Citations: 0

Authors: Li, Si; Yu, Chia-Hui; Wang, Yichuan; Babu, Yedurag

Abstract: The aim of this study is to propose an automatic and real-time social media anal ... Expand

Abstract: The aim of this study is to propose an automatic and real-time social media analytics framework with interactive data visualizations to support effective exploration of knowledge about adverse drug reaction (ADR) surveillance. This proposed framework has been prototypically implemented on the basis of social media data. A longitudinal diabetes patient online community data (AskaPatient.com) as well as FDA Adverse Event Reporting Systems (FAERS) data as a benchmark were used to evaluate our proposed approach’s performance. Based on the results, our approach significantly increases the precision and accuracy for ADR extraction. The number of ADR cases, the time when the ADRs occurred, and the rating of Glucophage have been visualized that resulted by mining a collection of 870 ADRs posted in Askapatents.com over a certain time period (from 2001 to 2015). The results have important implications for pharmaceutical companies and hospitals wishing to monitor ADRs of medicines. Collapse

Topics: social media analytics social media data visualization interactive data visualization electronic health

Methods: machine learning exploratory data analysis longitudinal research naïve bayes case study

User choice of interactive data visualization format: The effects of cognitive style and spatial ability

2019 | Decision Support Systems | Citations: 5

Authors: Luo, Wenhong

Abstract: Interactive data visualization has become an important tool for business executi ... Expand

Abstract: Interactive data visualization has become an important tool for business executives in monitoring key performance indicators, analyzing business process data, and managing individual and organizational performances. One of the key features in interactive data visualization is that users have direct control over how information is presented during the decision-making process. Few studies have examined how users choose visualization formats in such an environment. In this study, we investigate how cognitive style, spatial ability, and task difficulty affect the preference and choice of visualization formats and how the chosen visualization format in turn affects decision accuracy and confidence. Our results show that the choice of visualization format is affected by a user's cognitive style and task difficulty. While decision performances in accuracy and confidence are affected by task difficulty, spatial ability, and visualization format, they are largely not affected by the interactions between visualization format and cognitive style, spatial ability, and task difficulty. Collapse

Topics: task characteristic data visualization task productivity problem solving business intelligence

Methods: exploratory data analysis experiment survey design statistical hypothesis test parametric test

Theories: cognitive fit theory dual-coding theory

Media professionals’ opinions about interactive visualizations of political polarization during Brazilian presidential campaigns on Twitter

2017 | HICSS | Citations: 0

Authors: Santos, Caroline Q; Manssour, Isabel H; Silveira, Milene S; Ruiz, Duncan D A; Trasel, Marcelo; Barros, Rodrigo C

Abstract: Interactive data visualization techniques are an important way to obtain informa ... Expand

Abstract: Interactive data visualization techniques are an important way to obtain information from large datasets. Data journalism is an emerging area that strongly makes use of such techniques. In this work we investigate the relationship between journalists (and media professionals) in their job routine and data visualization, with the main goal of understanding if these professionals know and use data visualization tools in their job context, as well as if they consider these resources to be important. For this, we present the results of a survey made with journalists and media professionals to analyze how interactive visualizations could help them to get insight or knowledge of such data, and if their use may improve and support these professionals’ activities. The results indicate that visualization and data analysis tools are still not easily accessible by those professionals, and therefore still less influential than they could be. However, most participants considered data visualization a valuable resource in their news production routines. As a contribution, we also identified positive points and understanding gaps of visualizations, as well as the perception of journalists and media professionals about getting information from data visualization. Collapse

Topics: Twitter data visualization social media social network database system

Methods: survey exploratory data analysis personal interview naïve bayes sentiment analysis

Sentiment Analysis Methodology for Social Web Intelligence

2015 | Americas Conference on Information Systems | Citations: 0

Authors: Bjurstrom, Sean; Plachkinova, Miloslava

Abstract: The rapid development and use of Web 2.0 and social media have stimulated user-g ... Expand

Abstract: The rapid development and use of Web 2.0 and social media have stimulated user-generated big data. Social and opinion-based websites offer a lot of opportunities for organizations to gather information on their customers’ behavior and needs and sentiment analysis emerged as a technique for analyzing unstructured data. However, there are still problems related to its adoption by businesses. The current study proposes a comparison of two similar methodologies for sentiment analysis which can be utilized by any type of organization. The methods were demonstrated using secondary data from Cars.com, Yelp, and Twitter. The presented methodology provides a detailed guide for practitioners to incorporate sentiment analysis in business operations. It does not require significant investments in infrastructure or specialized software, which makes it much more affordable. Current knowledge on web social intelligence is extended by demonstrating how incorporating a number of methods and tools can be successful for analyzing unstructured text. Collapse

Topics: big data social media website Twitter decision making

Methods: sentiment analysis exploratory data analysis case study longitudinal research