Search

The age of big data drives the need for emerging technologies to enable scalable analytics on massive, rapidly generated, and varied data. It requires "data scientists" with deep knowledge of managing the six Vs of big data: volume, velocity, variety, volatility, veracity, and value. As a result of this trend, new analytical tools are being taught in business analytics (BA) programs to foster students' development of this critical competency. In the era of big data, social media analytics has become an increasingly important topic. In this article, we present a social media analytics exercise that can be easily added to any analytics course or to any course in which students gain exposure to social media and big data technology. The case scenario is built upon using flu activity data on Twitter to extend the monitoring of flu outbreaks. Our analytics framework comprises temporal, spatial, and text mining. We demonstrate the use of IBM InfoSphere BigInsights -- a Hadoop-based platform -- for implementing our framework. The exercise guides students through a big data social media analytics journey that enables them to understand different aspects of social media as the main source for big data and develop skills in this emerging area. Our framework and the exercises are general enough to be used even if the instructor opts to use a different technology stack. The described approach was used in a class at a large university. At the end of the exercise, 27 students majoring in business analytics participated in a survey and expressed satisfaction with their learning process.

In this article, we present an experiential perspective on how a big data analytics course was designed and delivered to students at a major Midwestern university. In reference to the "MSIS 2006 Model Curriculum," we designed this course as a level 2 course, with prerequisites in databases, computer programming, statistics, and data mining. Students in the class were mostly seniors or at the graduate level, and had a strong technical and quantitative background. We include details of concepts covered in the course, as well as summaries of four major sample course assignments used. Some of the concepts covered include large-scale data collection and management using the Hadoop ecosystem, stream mining, visual analytics, and social network analytics. Besides Hadoop, the course also introduced various IBM and Teradata big data tools. We show how the course modules align with the intended learning goals and course objectives. A post-course survey indicated that the structure and organization of the course helped students clearly and concisely assimilate the course content.

A reliable and accurate estimate of the expected hospital length of stay (LOS) of a patient is important to patients, medical providers, and insurance companies. Predicting hospital Length of Stay (LOS) is a complex and ill-structured problem, driven by many factors such as a patient's individual characteristics, treatment plans, and disease-interactions. In this paper, we develop a novel model to predict the expected LOS at the time of admission by combining network science and deep learning. We propose a two-dimensional construct of latent comorbidities comprising historical and probable comorbidities that a patient does not currently manifest but could likely develop during the course of hospital stay. The probable comorbidities are derived from a network comprising relationships among diseases observed in 3.2 million patient records in hundreds of US hospitals. We employ this construct of latent comorbidities in deep learning models to predict patients' LOS using almost 10 million other patient visits belonging to various disease categories. Implementing these models and analyses required a high-performance computing (Big Data) facility. The average mean absolute percent error of our models across all categories of diseases was 29.8%, which is the best in the current state-of-the-art. Our primary contribution is in developing a generalizable method to create a predictor construct for recognizing underlying relationships through network analyses, which can then be used in a deep learning model to predict an exogenous dependent variable. [ABSTRACT FROM AUTHOR]

Digital native and digital immigrant user types characterise the differences between those who grew up in a world of ubiquitous information systems and those who pre-date it. The rise in computer-mediated communication (CMC) technologies such as emails, text messages, and social media creates more opportunities for interruptions. Researchers have explored the impact of growing up in a world of technology. However, little research has explored the potential differences between digital natives and digital immigrants handling technology interruptions. This paper uses a qualitative approach to examine the extent to which individuals handle CMC interruptions differently based on the role of technology and its level of pervasiveness in the environment in which they grew up. Thirty-three interviews were conducted with individuals from different population groups based on age and gender to understand how technology interruptions are handled in work and social life. The findings support existing pre- and post-technology interruption frameworks and explain how and why generational differences exist in the process of CMC interruption handling. The theoretical and practical implications are discussed. [ABSTRACT FROM AUTHOR]

Within the context of the telecom industry, this teaching case is an active learning analytics exercise to help students build hands-on expertise on how to utilize Big Data to solve a business problem. Particularly, the case utilizes an analytics method to help develop a customer retention strategy to mitigate against an increasing customer churn problem in a telecom company. Traditionally, the forecast of customer churn uses various demographic and cell phone usage data. Big Data techniques permit a much finer granularity in the prediction of churn by analyzing specific activities a customer undertakes before churning. The authors help students to understand how data from customer interactions with the company through multiple channels can be combined to create a "session." Subsequently, the authors demonstrate the use of effective visualization to identify the most relevant paths to customer churn. The Teradata Aster Big Data platform is used in developing this case study.

Analytics researchers are widely using network analysis as a part of their methodology. In this review paper, we discuss different network concepts while summarizing some studies conducted using descriptive, predictive, and prescriptive analytics approaches. These applications illustrate the value of incorporating network properties of a phenomenon in better understanding the problem, prediction, and optimization of an outcome of interest, especially in the health domain. [ABSTRACT FROM AUTHOR]