|
|
Characteristics of Information Systems 3.5.1 Passive and Interactive Information Systems An information system is a system that provides information according to a user’s requests. Passive Information Systems Passive information systems are systems that will answer queries based on the data that is held within them, but the data is not altered. A simple example would be an electronic encyclopaedia where queries can be used to search for data and much valuable information can be learned, but the user is not allowed to alter the data. Another example would be the student file in a school that can be accessed by members of the teaching staff to find out where a student is at a particular time of day, or to look up their telephone number in order to contact the parents. The database of information is a valuable resource but it is not possible for an ordinary teacher to alter it. Interactive Information Systems An interactive system is one that data can be entered for processing which may alter the contents of the database. An example would be the school secretary updating the attendance record of a pupil in the pupil file. In commerce, a stock control system in a supermarket is an interactive information system because it not only gives information like the price and the description of the goods for the till receipt (passive), but also updates the number in stock immediately (interactive) so that when the next item is sold the number in stock has already been altered. 3.5.2 Management Information Systems A management information system is software that allows the managers within a company to access and analyse data. The management within a company, or any other organisation, is usually working on two levels. These levels may be different levels of management or may be the same people but with two different tasks. One is the day to day management of the organisation, keeping track of invoices, ensuring that the business has enough stock, arranging for the workers to be paid… The other is the strategic management, making decisions about what items of stock to keep, what new items to stock, what to discontinue… The first of these management tasks is condition driven. The information system can be used to trigger decisions that have to be made. A certain item is reported to be low in stock by the system requiring a decision to be made as to whether to reorder and how many to reorder. The second of these management tasks is to look at the wider picture. To extract information from the system which can inform wider decisions like should an item be stocked, which areas of the business are doing well and which are not, which workers are having to do a lot of overtime which may point to the need to increase staff in that sector. A good MIS provides managers with appropriate information at the right time. It also presents information in the right amount of detail according to the level of management. Shop managers want to know details of sales of all the products in the shop that they manage. Area managers want details of total sales for each shop in the area for which they are responsible. Senior managers want details of sales and costs in areas, so that they can make decisions about, for example, whether or not to open a new shop or close an existing shop. If a new shop is to be opened, in which area should it be so as to be most profitable? The MIS needs to organize the data collected over a period of time and to turn the data into information for the management of the business. Information must be appropriate for the making of decisions at specified levels of management. Another example is the information available in a manufacturing business. The manager of the production line needs to know what orders are outstanding in order to decide the best order in which to fulfill them. The sales manager needs to know the overall picture of sales and orders not necessarily details of individual orders. The manager will also need to be able to answer questions such as Are sales of some products better than others? What effect has marketing had on sales? Has a marketing strategy in one area had a better effect than in another area? Summarising, a good MIS provides information that is relevant and accurate for its purpose, provides information that is sufficiently complete and reliable to instill confidence in the user, communicates the information to the right user in time for its purpose and targets its detail appropriately for the use to which it may be put, communicates the information to the user in such a way that it is understandable to the user. 3.5.3 Applications requiring Batch Processing/Rapid Response Batch Processing This work has been covered in Chapter 1. A batch processing system is used when the output does not have to be produced immediately. Other factors are that the application will tend to use a large amount of data, that processing will tend to be of the same type for each set of data and that human intervention is not necessary. In module 1 we saw that the perfect example of batch processing is the payroll system, and nothing has changed because we are in a different module. Rapid Response Processing In module 1 this was referred to as real time processing. Real time processing can be thought of as being used in process control where the results of the process are used to inform the next input. A good example would be the control of a robotic machine on the production line. The other example is in information systems where it becomes necessary to update the file of information before the next enquiry is dealt with. The classic example is the airline (or theatre) booking system. If a customer enquires whether there is a seat available on the flight, and in response to a positive reply decides to buy the ticket, then the number of tickets now available must be updated before the next person makes an enquiry, otherwise the second person may be sold the same seat. Similarly, the mail order company will need to change the number of items in stock before the next query arrives. In general terms, an examination question will expect the student to be able to decide which of these two types of processing are most appropriate in a given situation. It is not sensible to list a number of different applications for each type and to expect students to learn them. The application in a question may not be on the list and the student would not be demonstrating an understanding of the concepts even if it was. Far more sensible is to learn to recognise the characteristics of each of the types of processing in a given application. 3.5.4 Knowledge-based Systems A knowledge-based system (KBS) a system where all the expert human knowledge covering a particular topic is brought together and made available to the user through a computer system which uses the facts in its knowledge base by applying rules that may sensibly be applied to the knowledge. Human knowledge encompasses such a massive area of data that it is not reasonable to try to distill all human knowledge into one computer database. Because of this problem of volume, knowledge-based systems are restricted to a narrow area of knowledge. It may be geological patterns in the oil exploration industry, medical diagnosis in the doctor’s waiting room or personnel information in a company, all are well delineated areas of expertise where there are no grey areas of whether the data should be included or not. The access to this knowledge can be very haphazard unless the system follows certain rules. That is why all expert, or knowledge-based, systems include rule bases which determine how the data within the system relate to each other. There must be an algorithm, or set of algorithms, for determining how the rules in the rule base should be applied to the knowledge in the knowledge base. These algorithms are known as the inference engine which gives a method for searching and querying the knowledge and rule base. Finally there must be a user interface which allows the user to interrogate the knowledge base. The user interface will pass requests on to the inference engine which, in turn, interrogates the knowledge base. The user interface will prompt the user, often by producing a series of questions each of which will have a small number of possible answers, and each of which will have the effect of accessing smaller and smaller areas of the knowledge base. The user interface will also allow the user to ask for explanations of the reasoning behind the advice that has been given and the results will give conclusions with the probabilities of those conclusions being correct. 3.5.5 Types of Knowledge-based Systems Students should be able to discern three different types of knowledge-based system. Diagnostic The user interface gives a series of questions, each of which has a limited number of possible answers, each one of which leads to another question. Gradually, the amount of data in the knowledge base is reduced until there is only a small amount of relevant data which must provide the answer to the query. A good example is a medical diagnostic machine. This contains an enormous amount of knowledge about medical conditions. To read all of it would be a Herculean task. If the user interface asks whether the patient has a high temperature there are two possible answers which immediately divide the knowledge base into two parts. A series of similar questions will result in only a small amount of the knowledge base being relevant. The result may well be more than one possible diagnosis. In this case, probabilities will be assigned to each diagnosis and a doctor has to finally reach a conclusion. Advice Giving An advice giving system is one that follows some process being done and then offers advice on how to proceed if something needs to be done or goes wrong. An example would be a stock control system which has knowledge of the stock levels of all the items and can offer advice as to whether further stock should be ordered, when to order, who to order from. Decision Making A decision making knowledge-based system is precisely that, a system which understands what is happening in a system and has been given enough rules to be able to make and carry out decisions without further intervention. A good example is the stock exchange share systems used by stock brokers. The computer system keeps a check on the values of shares constantly. The system has been given certain rules about what to do when share prices change. For example, the system may have been told to sell the shares owned by a customer if they fall below a certain value. The reason for this is that, if the shares are falling sharply, the instruction to sell when a particular value is reached can save the investor a lot of money. This has become necessary because shares are now traded 24 hours a day in stock markets all over the world, not just in the hours that, for instance, the London stock exchange is open. This means that if shares fell suddenly in the early hours of the morning an investor may be bankrupted before the stock market opens again. This has led to the automated buying and selling of shares by knowledge-based systems. There are dangers in such an approach. In the 1990s there was a big plunge in the stock markets all over the world caused by just such an automated approach. Some stock fell triggering some systems to automatically sell. This put more stock on the market which, in turn, brought the price down again, which triggered more machines to sell. It was very nearly a disaster caused by the fact that the knowledge bases were not big enough to see a wide enough picture. Because the detail was so narrow, the systems had no way of realising that they were causing the problem and had no rules for dealing with the situation even if they had. New rules have been written into the knowledge base to make such an occurrence far less likely to happen again in the future. 3.5.6 Business MIS and Knowledge-based Systems A management information system is a tool that can be used by a human being in order to make sense of and glean information from data in a database. A knowledge-based system is one that can do some of the reasoning that would otherwise have to be done by the human being, by following the rules in its rule base. Example Questions 1. Explain the difference between a passive information system and an interactive information system, giving an example of each. (4) A. -A passive system is one where the database is not changed -e.g. This book when it is installed on a network -An interactive system is one where it is normal for data that is either offered to it or which it produces will also have the effect of altering the data stored. -e.g. A system that controls the progress of a computer controlled robot vehicle around a factory. Notes: The amount of material that is covered in this section is so small that there are very few questions that can be asked, in fact this just about covers all that can be expected of a candidate. 2. A company payroll system uses a personnel file among others. a) Explain the difference between batch processing and rapid response processing. (2) b) Explain how batch processing and rapid response processing can both be sensibly used in the context of the personnel file being used to calculate the payroll. (4) A. a) -Batch processing is the processing of data by collecting the data and then processing it in a single batch. -Rapid response processing is processing where the result of the processing is needed immediately. b) -Payroll needs to be run once per week and.. -requires no human intervention, consequently perfect for batch processing. -Enquiries by employees need to be handled on a one off basis.. -with the result of the enquiry being made available to the employee immediately. Notes: The use of different operating systems at different times to perform tasks within the same application area provides a very popular question. 3. Describe what is meant by (i) a Management Information System (ii) a knowledge-based system. (4) A.(i) -An MIS is software that allows a manager to access.. -and analyse data. (ii) -A knowledge-based system is one that has all human knowledge on a restricted topic in the knowledge base.. -together with a number of rules for applying that knowledge.. -and an inference engine used to search the knowledge base. Notes: A very definition based question. The only other type of question that would be asked is one that asks for advantages and disadvantages of use of one of these systems in comparison with some other method of determining information. |