Use of Artificial Intelligence in Decision Support Systems
Posted by Winnie Melda on October 18th, 2018
There is a need for organizations and individual users to make good decisions. However, this is becoming increasingly hard. Organization managers have too many decisions that they are required to make, too much information at their disposal, and insufficient time to do any of that. Additionally, there is a change in objectives, and the diversity of things that need to be met has grown. Decision Support Systems (DSSs) can help in sorting and processing information and expand the ability to make good decisions in the light of these difficulties. Decision Support Systems (DSS) would consist of the following components (Checkland, 1981):
i. Data Management Component
ii. Model Management Component
iii. User Interface Management Component
iv. Decision Support Architecture
The user interface is the DSS component that links the user to the DSS, therefore, providing communication between the user and the DSS. The user interface should be designed properly because it is the only component which the user deals with. The main aim of this paper is briefly reviewing the use of artificial intelligence (AI) in decision support systems (DSS).
For the purpose of making good decisions, decision makers need to:
i. Comprehensively understand the problem
ii. Access relevant information differentiated from the irrelevant
iii. Have deep knowledge of using this information
iv. Have adequate time of applying the knowledge
Additionally, the decision makers must know the most effective and efficient strategies that need to be used while making decisions of various types. However, the task of making decisions is becoming increasingly difficult. The people are supposed to be making decisions have too much information at their disposal and again are required to make too many decisions in very short time.
The amount and sophistication of the information available to the managers are becoming more and more at a faster rate. With the advanced technologies such as artificial intelligence (AI) and remote sensing, organizations and individual users are collecting and storing information that was not collected a decade ago. The amount of both qualitative and quantitative written information has proliferated to the point where the pure amount of pertinent literature accessible to the decision makers has become virtually unmanageable. Additionally, there is persistence in the scarcity of human experts because people are shifting jobs while others are retiring and agencies and institutions restructure.
ROLE OF COMPUTERS IN DSS AND THEIR LIMITATIONS
Computers can help in the process of decision making. They are capable of sorting and processing information in various ways such as:
i. Providing the tools needed to capture while using both qualitative and quantitative knowledge.
ii. They can sort and process information faster than humans.
Therefore, with the availability of new computer tools, they can be implemented in the DSS thus improving their capabilities. Having a good and suitable DSS resource managers and decision makers can make good decisions with the availability of a wide range of information even if there is a need to make decisions very often or very rarely. However, developing useful programs and applications for DSS has been anything but easy.
Computers have been traditionally used to process information, process information, and manipulate models of various types. DSSs evolved responding to the need of making traditionally used management information systems and models more interactive and user-friendly thus integrating capabilities of data processing of the computers with the managerial judgment capabilities of a human. The main aim of early DSSs such as FORPLAN was resource management at a local or regional level. Increasingly, applications have continued to be integrated thus providing various functions capable of solving more sophisticated problems. Systems linked this way include databases, geographic information systems, data exchange systems required for the importation and exportation of data, simulation packages, and finally user presentation modules. The limitations of the above systems were (Ginzberg, Reitman, & Storh, 1982):
i. Difficulties in learning and using
ii. Difficulties of interpreting output from the answers
iii. They cannot be practically incorporated with intelligence that would make them easy to learn and to use.
ARTIFICIAL INTELLIGENCE (AI) AND DSS
The main aim of the field of artificial intelligence (AI) is adding humanlike features of behaviors to computers and systems. This aim includes making computers easier to use, developing effective and efficient communication between the computer and the user, and making the systems and computers deal with ideas and words, uncertain or incomplete information, and more sophisticated problems. AI was a formal discipline in as early as the 1950s and 1960s and its researchers looked to come up with ways of building a “general problem solver” an application founded on the idea that there is a basic human ability through which problems can be solved regardless of the subject matter. There is still discussion as to if there is such ability. In the modern times, it has become apparent that solving problems intelligently in humans and computers is based large amounts of knowledge in a specific domain. This has given rise to the AI field of expert systems (Negnevitsky, 2002).
Progress in the field of AI together with improvements in hardware and software for building AI systems has led to the rapid movement of the AI technology into the hands of people developing DSSs for decision makers and resource managers. AI enables people to encode knowledge and act upon information. This includes encoding output from various models intelligently. In case there are no qualitative models, qualitative aspects of the problem in question can be addressed using the approach. The main advantage of the AI in DSS is the application of systematic application of a large amount of specific, inferential, and context-dependent knowledge. Capturing of the best knowledge of the best people in an organization about to do with the available information is possible using modern DSS. This component of the process of decision making can be made more consistently accessible to the people who need it. Until the integration of AI into DSS this level of activity was the only domain of humans who are often overloaded with information, have insufficient time, and too much information to act upon (Carr, 1992).
Various tools and technologies from AI have added new capabilities to the DSS. Additionally, this has the development of DSS in organizations. Among these capabilities include development of mathematical software and hardware, new AI techniques, data mining, data warehouse and multidimensional databases (MDDB), online analytical processing (OLAP), intelligent agents, enterprise resource planning, telecommunication technologies like World Wide Web (WWW), corporate intranets, and the Internet. Therefore, more and more systems have incorporated domain knowledge, analysis systems, and modeling thus providing users with the capabilities of intelligent assistance. AI modules have been incorporated with the DSS to formulate problems and decision models. Additionally, they analyze and interpret the resultant output. Some DSS have also added AI modules thus replacing human judgments (Fedra, 1995).
The DSS based on AI include management of knowledge component for storing and managing a new class of emerging AI tools like the case-based learning and reasoning and machine learning. These tools have immense capabilities of obtaining knowledge from the previous data, decisions, and cases after which contributes to the creation of the DSS thus supporting repetitive, real-time, and complex decision making. Machine learning is the computational tools and methods of a computer system that learn from the past solutions, data, and observations, and consequently, modify its behaviors. The altering triggers this to the already stored knowledge (Henrion, Breese, & Horvitz, 1991).
The decision systems of the organizations cover the main part of the transaction processes of the large amounts of the processes at the level of tactical decision making. Various implementations of the information technology capabilities in various areas of human activities lead to gathering and availing a large amount of data. Thus consequently leads to rapid growth of the volume of internal and external databases. The problem here is utilizing the data they contain. It is possible to get the new information from the large volumes of incompatible databases, but the process is very inefficient. This calls for DSS which needs to incorporate with AI. This also supports various strong examples from many industries as the paper has shown. Therefore, as much as organizations are going to implement DSS, there is a need to incorporate them AI to improve the efficiency and effectiveness.
Carr, C. (1992). Performance support systems: A new horizon for expert systems. AI Expert, 7 (5), 44-49.
Checkland, P. B. (1981). Systems Thinking, Systems Practice. Chichester: John Wiley.
Fedra, K. (1995). Decision support for natural resources management: Models, GIS, and expert systems. AI Applications, 9 (3), 3-19.
Ginzberg, M. J., Reitman, W., & Storh, E. A. (1982). Decision Support Systems. Amsterdam: North Holland Publishers.
Henrion, M., Breese, J. S., & Horvitz, E. J. (1991). Decision Analysis and Expert Systems. AI Magazine, 12 (4), 64-91.
Negnevitsky, M. (2002). Artificial Intelligence: A Guide to Intelligent Systems. London: Pearson Education.
Sherry Roberts is the author of this paper. A senior editor at Melda Research in best nursing writing services if you need a similar paper you can place your order for custom nursing papers.
Like it? Share it!
About the AuthorWinnie Melda
Joined: December 7th, 2017
Articles Posted: 364
More by this author