15 Decision Support Systems

Dr. Ashish Saihjpal

epgp books

 

 

 

1. Learning Outcome:

 

After completing this module the students will be able to:

  • Understand how MIS can be applied for Decision Support Systems (DSS)
  • Understand fundamentals of Decision Support Systems (DSS)
  • Understand the basic types of Decision Support Systems
  • List various advantages and disadvantages concerning DSS
  • Get an overview of various successful DSS implementations
  • Understand the use of DSS from an analytical modeling perspective
  1. Introduction

 

Decision support systems are application programs that can be fed with business related data and yield information that supports decision making. It helps to process voluminous data to generate business useful information. This otherwise can be a cumbersome task. Decision support systems take input from other information systems in an enterprise. Some of these may include accounting systems, human resource information systems, office automation suite, transaction records etc. These decision support systems fetch data from data aggregation techniques, summarize information, prepare trends and patterns and yield logical analysis. Such a system uses numerous analytical modeling techniques.

 

Data is pooled in from multiple resources such as transactions, documentation, customer feedback, daily reports etc (Exhibit 1). DSS integrates with such sources to access raw data for analysis. DSS support numerous statistical and research oriented techniques like decision theory, probability, operation research, artificial intelligence etc.

 

Proper integration of DSS increases the efficiency and productivity of business. This helps enterprises to seek an edge over peers in the industry by analyzing the available resources, logistics and investments better. However, decision support systems are limited to providing technical and analytical assistance. They cannot be replaced by human experts.

 

Let us assume the scenario of an industrial plant. The problem at hand is to find the possible effect of introducing the third shift in a factory? One might expect that this would increase the plant’s output by roughly 50 percent. Such decisions are liable to have financial implications as factors such as additional wages, maintenance, machine wear and tear, raw material procurement, logistics and seasonal demand need to be considered. However, analyzing the impact on each of the variables and its effect on production is a hassled task. This is where decision support systems comes into play.

 

  1. Basic Components of a Decision Support System

 

The basic components of the decision support systems are exhibited in a block diagram below (Exhibit 3).

 

  1. The Data base: It is a collection of data from a number of the applications or data groups. The data base provides very easy access for a numerous business processes. Maintaining and updating a database requires the expertise of a database expert. The otherwise unstructured information is organized and grouped in a manner, that it can be easily managed and updated. Running a database query to fetch details becomes easier as data is accessible and easy to locate.
  2. The Model – A model is a representation of the system in a particular language that the computer understands. To interpret a model one needs translation of the computer programming language to a user friendly format. There can be a physical model, a mathematical model or analytical model. Analytical modeling refers to the use of data groups, algorithms and mathematical techniques and studies the behavior on data. Various types of the models can be categorized as the follows:

Behavioral Model – The underlying focus is on studying and understanding the different behavior/trends amongst the variables. Examples of such a model can be trend analysis (Exhibit 4), correlation, regression etc.

 

  1. Management science model – These models are based on principles of science and management. Some of the application areas include decision theory, accounting, capital budgeting, inventory analysis (Exhibit 5) etc.

C. Operations research model – This kind of a model is based on the use of mathematical formulae. It can be applied to solve a number of routine problems. Some of the examples include ABC analysis, Pareto analysis (Exhibit 6) and mathematical programming techniques.

3. The Decision support software– This software allows interaction between the user of the system and the DSS. Data from external sources is fed into the DSS which helps in analysis, storage and retrieval of data from the central database for business use. The DSS has a friendly graphic interface which makes it easy to use (Refer Exhibit 7).

4. User interface – It refers to an interactive screen that makes it easier for a non expert to use the software. This not only makes the software self explanatory but helps to make the analysis
simpler.

5. The Hardware configuration – It is the size of the database and the scale of operations of a firm that determine the hardware configuration. The requirement of the data base management software package can be decided accordingly. The hardware should be expandable and scalable as per growing business needs. Other features that determine the hardware configuration is the type of analysis expected out of a DSS, the model used and reporting engine functionality.

 

  1. Types of Decision Support Systems

 

There are a number of Decision Support Systems. These can be categorized into five types as shown in figure1and explained below:

 

  • Communication-driven DSS – These systems are designed to enable interaction and collaborative techniques, where more than one individual needs to work together. For example, a meeting conducted over video conferencing by the sales department to agree on software customization with agreement from the product development team. The product team in this context could be based out of a different state, country or continent.
  • Document-driven DSS – They assist in document archival and search from a central database by running a database query. These systems are integrated with data mining and knowledge management tools to pull out the appropriate documentation set for business related activities e.g., a system update, product related documentation, specification sheet etc.

 

  • Data-driven DSS – These systems are targeted at mid level of the organization and operational managers who seek answers to queries run on a database. These teams undertake operational activities for business analysis. They may process queries over the web or on a client server link. Exhibit 8 showcases a data driven DSS with different stages of the data processing cycle.
  • knowledge-driven DSS –It involves the various stakeholders who are responsible for taking These can be people within or outside the organization. It may include end consumers, suppliers or vendors.
  • Model-driven DSS – These systems are centered on the use of data and analytical modeling They help to make statistical, inferential or simulation based decisions. The infrastructure can be laid out using stand alone systems or client server architecture. E.g., what if Analysis, computer aided design, computer aided manufacturing etc.
  1. Advantages and Disadvantages of using Decision Support Systems

 

The major advantages or benefits of implementing a DSS can be summarized as follows:

  • A DSS enables the solution of complex problems that ordinarily cannot be solved by other computerized approaches like a spreadsheet.
  • An easy to understand user interface allows the user to carry complex data processing and analysis and interpret results easily.
  • DSS requires much lesser time than a team of human experts to analyze large volumes of data or process the same task repetitively.
  • Business decisions are interdependent among different departments. The use of decision support tools leads to better team management as members across different teams can achieve results that link cross functional departments.
  • Decision support systems can be used for multiple applications on a day to day routine. This makes a user well versed with the usage and helps to resolve problems more critically.

 

Decision Support Systems are intended to improve the accuracy, timeliness, quality and overall effectiveness of a specific decision or a set of related decisions. However, viewing from another perspective, they do come with some disadvantages such as:

 

Disadvantages

  • It increases dependency on the system software for analysis and support which thereby reduces problem solving ability in workers.
  • Reduction in efficiency may occur because of information overload.
  • It shifts the responsibility and blame on the machine for errors that may incur. This leaves employees to feel they are now left with clerical and administrative work.

 

5.1  Case of Intelligent Decision Support System in Clinical Diagnosis

The Intelligent Decision Support Systems (IDSS) are very competent counterparts of clinical decision support as they make extensive use of artificial intelligence. These systems can rightly be called human consultants since they acquire and gather evidence, perform diagnostic checks and make suggestions and course of action. Features of the clinical decision support system are highlighted in Exhibit 10. The IDSS when integrated with the knowledge based systems can perform analysis based on patient data, medical history, family history and lab reports concurrently.

 

In a particular scenario, Dr Smith has an Electronic Health Record System (EHR). His patient, Jade uses a similar version called the Patient Health Record System (PHR). Jade, on account of diabetes takes a combination of Metformin, Aspirin and Lipitor. Her average BP is 108/55 mm Hg and heart rate 60 beats/minute. She self monitors her BP, blood glucose four times a day. She feeds this data into the PHR which is integrated with the EHR of Dr Smith. The data sets are stored in agreement with the medical care guidelines. Data can be fetched and derived to undertake any study of the patient’s medical history concerning any ailment that needs monitoring.

 

One morning, the doctor observes consistent BP and heart rate but mild systolic murmur. Usually, he recommends a blood thinner in this case. But, the same cannot be generalized for Jade as it may cause an overdose if she takes aspirin.

 

The intelligent system closely monitors the pattern of BP and Heart Rate and sends an alert to the doctor. His intervention at this stage is necessary. Had jade assumed the condition to be normal as both parameters were within prescribed limits, she would have still suffered bleeding. Since, the IDSS was integrated with the EHR it triggered the drug-drug interaction alert and prevented the condition to go out of hands. This precision is of utmost important in medical conditions where the slightest negligence can prove fatal.

  1. Role of Decision Support System in MIS

 

An organization functions by taking decisions at each step. What differs is the nature of the task, the risk involved and the investment needed. While operational activities need manual intervention but they can be automated with use of information systems. At the mid management and strategic levels the decisions involve higher stakes and risk. Hence, the use of decision support systems becomes important.

 

Decision support system involves the packages which help the managers to take accurate and timely decisions. DSS use data from the general management information system and they are used by a manager or a decision maker for decision support. Decision support system utilize a number of pre-programmed techniques or mathematical models. These systems are used for testing new alternatives, training and learning.

 

The management information system would become more useful if the ability to make decisions is user independent. The system works under a given set of assumptions and limiting factors. If the system needs to look beyond the scope of these assumptions, decision support system alerts the decision makers that action is called for in the situation. The decision support system plays an imperative role in the management information system as a support to decision making.

 

 

6.1 Using Decision Support Systems

 

Decision Support Systems work on the principle of analytical modeling. This is different from the principle of management information systems. MIS tend to fetch information relevant to business. On the other hand, DSS list possible alternatives under given conditions. The four basic analytical modeling activities involved in using a DSS are represented in figure 3.

  • what – if analysis: The user makes changes to variables or relationships among variables, and observes the resulting changes in the values of other variables e.g., What if we cut advertising by 18 percent? What would happen to sales? This type of analysis would be repeated until the manager was satisfied with what the results revealed about effects of various possible outcomes.
  • sensitivity analysis: Sensitivity analysis involves the change in only one type of value. This change is performed repeatedly and the corresponding effect on other variables is monitored. Decision makers use the technique of sensitivity analysis when there is un-certainty of the chosen assumptions in estimating the value of certain key variables.
  • Goal – Seeking analysis: It helps in back calculating the values. The aim is to obtain that particular input that would result from a certain output. e.g. what quantity of products should be sold by each sales executive of till the sales revenue reaches INR 1, 00,000? The Goal seek analysis is an inbuilt feature of MS Excel package. This is easy to use. A snapshot is represented in figure 4. It suggests, what marks should be scored in Test 3 to achieve a score of 70%.
  • Optimization analysis: Fundamentally, the word optimization refers to minimizing or maximizing certain attributes that shall have a positive impact on the organization (refer exhibit 11, the change before and after optimal thinking). It can help in profit maximization and cost minimization. In context of decision support, the aim here is to find the target value for more than one variable. Keeping in mind the limiting factors or constraints, the values for each of the variable is changed iteratively till best value is identified. e.g., what is the best amount of advertising to have, given our budget and choice of media? What is the highest profit that can be earned while auctioning of a house?
  1. Summary

 

DSS refer to interactive software systems that translate raw data, documents, customer profiles and other relevant information using various techniques and business models to solutions that assist in decision making. DSS have found applications across all functional areas and different industries. A step ahead, the Intelligent Decision Support Systems integrate the use of artificial intelligence and work like consultants. These help in optimization, efficiency, business competence and ease in problem solving. In order that profitable solutions may be identified, more alternatives may need to be explored and some decisions may need to be automated. Web based technologies have further enhanced the capabilities of decision support systems. The tools and techniques for analysis in a decision support system application package are integrated and programmed by humans. An unaided subject expert is equally competent to take decisions but may tend to overlook some areas if large volume of data is to be analyzed. This is where an automated system assists to fragment and make the tasks simpler. Hence, while DSS cannot replace the human intervention, diligent interfacing can help build competent solutions.

you can view video on Decision Support Systems

Learn More:-

  • Druzdzel J Marek, Flynn Roger R, “Decision Support Systems”, Decision Systems Laboratory School of Information Sciences and Intelligent Systems Program, University of Pittsburgh.
  • Laudon Kenneth C, Laudon Jane P, Management Information Systems, Managing the Digital Firm, Pearson Education South Asia, 2013.
  • O’Brien A James, Marakas M George, Behl Ramesh, “Management Information Systems.” 9th Edition, Tata Mc Graw Hill Education Pvt Ltd.
  • Thapliyal MP, Kautish Satish, “Concept of Decision Support Systems in relation with Knowledge Management – Fundamentals, theories, frameworks and practices”, International Journal of Application or Innovation in Engineering & Management, Vol 1, Issue 2, 2012, ISSN 2319 – 4847.