|
Thought Leader Interview
Clyde Holsapple: Big Picture about DSS
Rosenthal Endowed Chair in MIS/DSIS
at Gatton College of Business, University of Kentucky |
Preface
Dan Power spoke with Clyde Holsapple at AMCIS 2004 in New York, NY and arranged an email interview. Professor Holsapple responded to seven open-ended questions.
Q1: How do you define the concept "Decision Support System"?
Holsapple's Response: I define a decision support system as a system that supplements, complements, or amplifies the knowledge resources and/or knowledge processing capabilities of a user engaged in making a decision.
We ordinarily think of a DSS as being a technological system, but should also recognize the existence of human decision support systems (HDSSs); that is, the “system” can be a person or social entity. When I use the DSS term, I mean computer-based systems, including those that support individual or multiparticipant decision makers. The user of a DSS is typically thought of as being either a person (individual) or a social entity having multiple participating individuals such as a group, hierarchic team, or complex organization. However, a user conceivably can be an individual or multiparticipant computer-based system.
The above definition recognizes decision making as a knowledge-intensive process. That is, the decision maker is viewed as a knowledge processor that, in the course of a decision making episode, can acquire knowledge, select previously-stored knowledge, generate knowledge, assimilate knowledge into storage, and emit knowledge. A DSS can help the decision maker with one or more of these kinds of activities as the episode unfolds. The outcome of the episode is new knowledge characterizing a commitment to some action, and possibly knowledge by-products as well. The configuration of knowledge processing that occurs in a particular decisional episode is influenced by resources available to the decision maker, by various managerial factors that impinge on or facilitate knowledge processing, and by environmental factors (e.g., socio-economic, legal, regulatory, market).
This perspective on DSS does not define decision support systems in terms of a particular architecture. While any particular DSS certainly has an architecture, it is not the architecture that makes it a decision support system, but rather its ability to hold and process knowledge that is of benefit to the decision maker as he/she drives toward creating knowledge about what to do. Nevertheless, the study of DSS architectures is certainly important; of particular interest is a generic architecture that can be specialized in a variety of ways to understand various classes of decision support systems.
DSSs are often said to be applicable to supporting unstructured or semi-structured decisions. While these may be the most challenging DSSs to develop or may have the greatest potential payoffs, computer-based technology can also be valuable in supporting structured decisions (e.g., able to cope with some of the complex or voluminous knowledge processing involved in a structured decision more rapidly or more accurately than would otherwise be the case).
As a footnote, the term “knowledge” as used here is in the sense of Allen Newell’s notion of that which is conveyed in usable representations. The representations can be digital, audio, visual, symbolic, material, mental, behavioral, and so forth. There is a tendency in the knowledge management literature to emphasize a distinction between tacit and explicit modes of knowledge. However, we should take into account the other attribute dimensions. One of these other dimensions of fundamental importance to decision making is knowledge type, particularly the distinctions between descriptive, procedural, and reasoning knowledge. Each of these three plays a significant role in decision making. Each is amenable to computer-based representation and processing, and therefore a candidate for DSS treatment. For each of these knowledge types, there are gradations in the degree of usefulness. The distinction that people often make between “data” and “information” is an example of different utility gradations in descriptive knowledge.
Q2: How did you get interested in this topic and what keeps you interested?
Holsapple's Response: In 1974, I had the good fortune of taking a Purdue University graduate course dealing with decision support systems taught by Andy Whinston. This may have been one of the earliest such courses to be offered in a business school. Andy sparked my interest in the DSS topic through the great vision and multidisciplinary approach he took to exploring the subject.
Those were very formative days in the history of DSS. It was clear that there was much to be done in terms of shaping and contributing to this new kind of business computing system. The DSS genre was emerging to give computers a fundamentally new role in business, beyond what was typical in the earlier genres of data processing systems and management information systems.
Completed in 1977, my doctoral dissertation identified some possibilities and advanced certain directions for the DSS field. This included the integration of AI techniques, database techniques, and solver techniques into a single system for supporting decision makers. It also outlined and advocated a knowledge-oriented view of decision support. Collaborating with Andy Whinston and Bob Bonczek over the next few years yielded a series of papers building on, refining, and extending the earlier work in a variety of directions. Our 1981 Foundations of Decision Support Systems book brought much of this work together into a single forum.
Writing in the foreword of that book, Herbert Simon characterized it as “pioneering work” that would bring about its own obsolescence. This latter point has long served as a spur that keeps me interested in the topic of decision support systems, in the senses of striving to make a continuing stream of contributions emanating from and improving on that early work, monitoring the advances that other researchers make in the DSS arena, and monitoring the advances made by DSS practitioners and tool vendors.
Q3: What are the new challenges and frontiers?
Holsapple's Response: New challenges and frontiers in the DSS arena include investigations of the evolving position of DSSs in the knowledge work of organizations (including better understanding of knowledge work itself), the unification of DSS technology with systems for organizational computing (e.g., enterprise systems, workflow systems, collaborative systems), the effective development and administration of diverse DSSs in an organization, the ways and degrees to which alternative DSSs can positively impact decision making in various situations (e.g., crisis situations, inter-organizational situations), new DSS applications, the greater incorporation of emerging technologies into DSSs (e.g., GPS, machine learning, multimedia, wireless), and the prospective roles of DSSs in such megatrends as pervasive computing, mass customization, globalization, continuous learning, and virtual organizations.
Q4: What do we really know about how computerized decision support impacts decision making?
Holsapple's Response: Decision support systems, in their many variations and forms, are in widespread use for supporting decisions large and small. Generally, we know that DSSs can and do have positive impacts on decision making. They can directly increase the productivity of decision-making processes, the quality of decision-making outcomes, and the satisfaction of decision makers. I suspect they can indirectly impact an organization’s locus of decision making, its agility, its innovativeness, and its reputation – all of which can drive competitiveness. Yet, this question deserves more research to better understand situational factors, user factors, and system factors that color the extent and nature of a particular decision support system’s impacts on decision making.
Q5: Have you encountered any "downside" to computerized DSS?
Holsapple's Response: One possible downside to DSSs is related to over-dependence in which users begin to treat them as decision making systems rather than decision support systems. Another is related to inadequate control over the development and maintenance of a DSS, resulting in a system whose knowledge and/or knowledge processing is faulty. There is also a possible downside if inconsistencies among systems in an organization’s portfolio of DSSs are allowed to flourish.
Q6: Why are "retrieval only" or data-driven DSS so popular?
Holsapple's Response: We can hypothesize several reasons. First, access to relevant, pre-existing descriptive knowledge is sufficient support for many decision making circumstances. Second, such DSSs tend to require less effort to operate and to understand results. Third, available software tools may make these DSSs relatively straightforward to implement, suggesting that they proliferate more rapidly than other kinds of DSSs. Fourth, the population of potential users for more complex or sophisticated DSSs may be relatively modest at this time. Fifth, such DSSs may tend to have a faster or more significant payback.
Q7: How is knowledge management related to DSS?
Holsapple's Response: Decision making is one aspect of knowledge management. Many knowledge management episodes that occur in an organization have decisions as their purpose and outcome. Other knowledge management episodes have other purposes and outcomes (e.g., research, communication). It follows that the notions of decision support and DSSs fall within the domain of knowledge management (KM). This suggests that an understanding of knowledge management would be useful (if not essential) for DSS researchers and educators. It also suggests that DSS researchers can contribute to a better understanding of KM. In short, DSSs comprise a class of knowledge technology. Care must be taken to realize that KM involves many non-technological matters as well as knowledge technology; these matters may very well help DSS researchers and practitioners in understanding contexts for DSS development and usage.
About Clyde HolsappleProfessor Holsapple holds the Rosenthal Endowed Chair in Management Information Systems and is Professor of Decision Science and Information Systems at the University of Kentucky. His research focuses on supporting knowledge work, particularly in decision-making contexts. He has authored over 100 research articles in journals such as Decision Sciences, Operations Research, Decision Support Systems, Journal of Management Information Systems, Group Decision and Negotiation, Journal of Operations Management, Organization Science, Communications of the ACM, Journal of American Society for Information Science and Technology, Knowledge and Process Management, Journal of Knowledge Management, and IEEE journals. His many books include Foundations of Decision Support Systems, Decision Support Systems: A Knowledge-Based Approach, and the 2-volume Handbook on Knowledge Management, a basic reference work. He has served as Editor of Journal of Organizational Computing and Electronic Commerce, Area Editor of Decision Support Systems and INFORMS Journal on Computing, and Associate Editor of Management Science.
CitationPower, D., "Clyde Holsapple Interview: Big Picture about DSS", DSSResources.COM, 02/11/2005.
Clyde Holsapple provided permission to publish this interview at DSSResources.COM on January 29, 2005.
|