The origins of decision support systems (DSSs) can be traced back to the 1970s, when the promise of enhanced managerial decision making through the use of models and the data processing capabilities of computers were first touted. Some would argue that this history goes back even further. However, it is not our intention in this issue of Cutter Business Technology Journal (CBTJ) to review the entire DSS literature, as many others have done a fine job of this already.1
Over the last 50-plus years, DSSs have been applied in numerous settings — whether it be allocating resources during an emergency, predicting crime, improving medical outcomes in a hospital setting, or optimizing a company’s supply chain. It could be argued that the evolution of DSSs is neither linear nor straightforward. Yet, despite fundamental changes to the underlying technology that drives DSSs, the discipline has thrived. Indeed, innovations in information and communications technology (ICT) have brought with them successive generations of DSSs. There will be many who argue that artificial intelligence (AI) and DSSs should not be grouped together; however, the expansion of DSSs and the influence of technically related disciplines has a historical precedent. Thus, we see DSSs playing a pivotal role as a reference discipline for advances in AI, big data, machine learning (ML), and data analytics.
The cutting edge of DSS development is that exciting intersection between DSSs and these new technologies. The human-centered lens of a DSS can provide a particularly advantageous perspective by promoting a people process and technology approach that goes beyond quantitative models alone. While AI research has attempted to replace the human decision maker, DSS researchers and practitioners have sought to assist the human decision maker. The next generation of DSSs must straddle these dual, often conflicting, purposes. Cloud-based storage and infrastructure-on-demand have removed many of the constraints traditionally associated with implementations, and advancements such as AI and ML have stretched the boundaries of what is possible. It is only natural, then, that such technology advances have engendered even greater optimism with regard to ICT-assisted decision making.
Our massive appetite for accumulating data is matched only by the inexorable advances in technologies that support the creation and storage of that data. However, it is through the analysis of such data that real value is derived. The positive impact of data-driven decision making on firm performance is well established. But big data, with its large volumes, high velocity, and variety, does not lend itself to the traditional analysis methods. A new breed of business analysts skilled in data science and Agile practices is crucial for the next wave of DSSs.
Lessons learned from data warehousing and online analytical processing systems can teach us how to glean insights from ever-increasing volumes of data. The application of quantitative data models and simulations to aid decision makers, as typified by model-driven DSSs, are more relevant than ever. Big data’s Achilles’ heel — the analysis of semistructured and unstructured data — is a well-trodden path in DSS research. Past research on dashboard design principles for data-driven DSSs can guide the data visualizations of tomorrow.
We can also learn from past mistakes unencumbered by the technical limitations of previous generations of DSSs. Previous work on group decision support systems, for example, quickly moved from systems that support colocated managers to systems that support geographically dispersed managers as communications technology improved. In the same way, the next wave of DSSs is evolving to incorporate data that is no longer restricted by geographic and organizational barriers. Social media activity; real-time, cloud-connected sensors; and virtual assistants are just some examples of the ever-growing sources of data. Identifying and prioritizing data sources of strategic value will be a crucial activity going forward.
For more than 50 years, the DSS discipline has evolved by incorporating the technological advances of the day. However, the longevity of DSSs is not due to technology alone. DSSs seek to leverage the insights and experience of users who interact with such systems. The discipline’s enduring success has always relied heavily on humans’ creativity and innovative capacity. As the field of DSSs has expanded to include other reference disciplines and technological advances, so too have the possibilities that such growth affords us.
Issues of values, trust, collaboration, and decision bias represent key challenges at the intersection between DSSs and AI. We contend that the practice of leveraging user insights and involving a wide variety of stakeholders in the collaborative design of DSSs is similarly prudent for AI and big data analytics projects. One only has to look at the example of the recent global financial crisis, and the subsequent rise of behavioral economics, to appreciate the limitations of quantitative models and the importance of unquantifiable insights. Deciding whom to rescue first in an emergency, for example, raises a dilemma that cannot be answered by technology alone. But technology can play a role by facilitating collaboration and communication between experts and leveraging past experience.
Hiring business analytics talent and leveraging skilled employees are key requirements for business intelligence (BI) success.2 Beyond technical and modeling skills, ethical and aesthetic concerns are often nonquantifiable and require participation from a broad range of stakeholders. Issues such as trust, motivation, and conflict are poorly understood in the realm of big data.3 Co-opting users into the design process may alleviate some of these concerns.
In This Issue
The challenge for this CBTJ issue was to accurately represent the diversity of research in the DSS arena while also giving a glimpse of the cutting-edge DSSs of tomorrow. As a starting point, Ciara Heavin and Daniel Power provide an overview of the design and development of modern BI and data-driven DSSs. They identify challenges and opportunities for managers and provide a sociotechnical view of DSSs by demonstrating practical guidelines for the people, process, and technology components of modern BI and data-driven DSSs.
The next two articles speak to the diversity of settings in applying DSSs. Both incorporate cautionary tales for practitioners tempered with practical solutions to address these concerns. Tom Butler and Leona O’Brien provide a timely perspective on AI in the financial industry. Their article provides a pragmatic perspective on the capabilities of AI and pours cold water on some of the hyperbolic claims made about AI and ML in the fintech and regtech space. The authors suggest a direction and guidelines for future research for AI to realize its potential in the financial services sector.
Next, Frederic Adam and Paidi O’Raghallaigh tackle the current healthcare crisis. They shine a light on the opportunities provided by medical decision support for clinicians and patients and identify a number of challenges to achieving connected health, which they define as “the use of technology-based solutions to deliver healthcare services remotely.” The article proposes a connected health blueprint that may well pave the way for future connected health systems.
The next two articles focus on emergency management (EM), providing guidelines for the use of scientific modeling technologies in disaster management and for contact tracing of airline passengers during a biological outbreak. The aim is to improve disaster mitigation, preparedness, response, and recovery, leading to better economic and human outcomes.
Theresa Jefferson and Gloria Phillips-Wren discuss modeling for disaster response. This is vital for practitioner/manager decision making to reduce the impact of a natural or man-made disaster. The authors examine the concept of technology embeddedness, noting that emergency managers must trust the technology and show a preference to use it prior to an actual disaster; the time to integrate technology into disaster recovery operations is not during a disaster. They explain how to effectively appropriate, integrate, and use modeling technologies for disaster response and, therefore, recovery.
In our final article, Michael Gleeson discusses how public health agencies and emergency managers can leverage the digitization of contact tracing of airline passengers at risk from a biological outbreak. He outlines the increased risk of infection and spread, facilitated by the increased numbers of airline passengers globally. A global framework to prepare for and respond to a biological threat, natural or otherwise, spread via air travel, can be achieved through the digitization of contact tracing using a collaborative approach among the airline industry, public health agencies, and EM practitioners. Identifying and locating at-risk passengers in a fast and efficient manner is paramount to limiting contagion spread.
Successive generations of DSSs have leveraged emerging technologies to enhance the breadth and scope of the domain. AI, ML, and data analytics will be the hallmarks of the next generation of technology-assisted DSSs. The articles in this issue highlight the technology that will drive the next generation of DSSs and illustrate the criticality of the human aspects of decision making. We advocate the importance of diverse stakeholder involvement in the analysis and design phases along with the use of technology to enhance collaboration and communication for distributed teams and improved decision making through training and simulation. Business analysts must augment traditional requirements-gathering techniques with data science skills. Project leaders must leverage their communication skills to champion new technologies and articulate the business value of AI-assisted decision support.
To learn about how new technologies will impact the traditional corporate environment, we need only look at users who experience needs similar to those of corporate users — but in extreme forms. Take, for example, cross-border or transboundary EM where multiple stakeholders often combine to make joint decisions irrespective of agency or geographic location. In such scenarios, strategic decision makers use complex data sets, risk models, and real-time sensor data while also leveraging the local insights of skilled decision makers. Decision making is enhanced by technology without ever being truly dependent on it. However, imagine a future where ML has the potential to enhance decision making by reducing cognitive bias and removing the stress associated with human decision making under duress.
From a societal perspective, dealing with large-scale issues such as connected healthcare and the prevention of pandemics will require a concerted effort from the private sector to cut across rivalries and administrative red tape. A more focused research approach is prescribed to meet the opportunities AI provides in the financial sector. There is also a huge opportunity for industry consortia to pave the way for the DSS/AI architecture of the future while also upholding ethical standards that adequately reflect wider societal values and morals. Each article in this issue provides practical insights for managers. In effect, we are not simply asking what the DSSs of the future can enable us to do. We are asking what we need to do to ensure the success of the next wave of DSSs.
Technology-Empowered Solutions: Redefining Decision Support
Explore the technologies that will drive the next generation of DSSs! This Cutter Business Technology Journal issue is available for purchase in the Cutter Bookstore.
1See, for example: Shim, J.P., et al. “Past, Present, and Future of Decision Support Technology.” Decision Support Systems, Vol. 33, No. 2, June 2002; and Power, Daniel J. “A Brief History of Decision Support Systems.” DSSResources.com, Version 2.8, 31 May 2003.
2Trieu, Van-Hau. “Getting Value from Business Intelligence Systems: A Review and Research Agenda.” Decision Support Systems, Vol. 93, January 2017.
3Phillips-Wren, Gloria, Lakshmi S. Iyer, Uday Kulkarni, and Thilini Ariyachandra. “Business Analytics in the Context of Big Data: A Roadmap for Research.” Communications of the AIS, Vol. 37, No. 23, 2015.