I've been officially in the IT field for about 15 years, and most of that time has been doing systems design. Very often, at work or when meeting new colleagues, disputes arise on professional topics, one of which is the answer to the question - "What is a system?" Everyone understands this concept in their own way, most often based on their experience and knowledge gained in the IT sphere; others interpret definitions taken from the Internet or textbooks. And most often these disputes do not lead to a consensus, and now I do not claim to be 100% accepting my reasoning.
In my opinion, the question of what a system is is one of the most important questions that any analyst must answer correctly. After all, we design systems, create new ones or change, develop existing ones, and if we do not understand what it is, what it consists of and how to work with them, then the quality of our work will be questionable.
Before moving on to the definition of the system, I will tell one story, namely the words of one of my teachers at the institute. In the last years of study, 4th or 5th year, I don’t remember exactly, he revealed to us the secret: "Systems do not exist." That is, after several years of teaching us the subtleties and methods of creating systems, he declares that systems do not exist. Looking at us with a smile on his face and after a short pause, he added that the system is not something that exists in our world: it is the ability to see the world around us in the form of some recurring dependencies. The ability of each person to see, to determine the consistency around us. And just as an artist sees the world in his own colors, so we see this systemic world in our own way. That is why I believe that our profession is the most creative.
We are creators, we paint the world with numbers.
This begs the question - what is needed to see the system, what distinguishes a simple vision from a systemic one? To begin with - understand what you are looking for, choose your definition that suits you, which you understand.
I will give my definition of the system, which I like best, and I use it everywhere as a basis. It is, in my opinion, the simplest and most understandable:
" A system is a set of objects and the rules of interaction between them"
To better understand what this is about, let's remember from our experience what systems we know: a system of equations from school; solar system; economic systems; social systems, etc. In all these systems, there are key objects that we consider: for example, equations, cosmic bodies of the solar system, people around us, as objects of the social system; and certain rules by which they interact within their systems: gravitational forces, social, interpersonal rules of communication, and so on. As was stated in the definition, all these systems are united by the presence of common elements - a set of objects and rules by which these objects are related to each other. Then it turns out that any person, defining the connection between two or more objects in the external world, creates a system.
Most likely, you are now thinking - what does the AIS have to do with it, how will it help us? Let's take a look at what an information system is. Let's take a look at our definition and refine it:
Informational (System is a set of objects and rules of interaction between them) => Information system is a set of informational objects and rules of informational interaction between them
It turns out that the difference between a physical system and an information system is in what objects are considered in these systems and what rules of interaction between them are distinguished.
Let us analyze what an information object is and what information interaction between these objects is.
In my opinion, an information object is an information model of a physical object stored as information, and it should point to a physical object
Let's look at an example: there are a dozen different physical objects around you, be it chairs, tables, a laptop, a bed, a sofa, and a couple of armchairs. And if you say: “Take that particular chair!”, Pointing to a specific physical object, then taking the right chair right away will not be difficult - in this case, you are working with a physical object. But when you break the attachment to a physical object, operating only by its name, you are already working with a model of a physical object - an information object : when you say "chair", you do not mean a specific chair, but simply highlight such a concept - an information concept, denoting an abstract information object with which you can continue to work - chair, table, armchair. If you say these words to a person who does not know the meanings of these models, he will not understand what physical objects are being discussed. You will need to explain what properties (parameters or attributes) you need to pay attention to in order to select the physical object you need from the given set of objects. It is these parameters of the information object that are its key element. An information object or, as I used to call it, an entity is a set of properties that distinguish the physical or information objects we need from the selected set of other physical or information objects.
To make it clearer, I will select 2 entities from the above example:
- chairs are objects defined by the presence of 4 legs, a base to sit on, and a backrest to lean on. The definition may not be accurate, but it is good for understanding.
- interior items that you can sit on are objects that have a flat support that you can sit on and that stand on the floor and rise to a height of at least 40 cm and not more than 1.5 meters.
It can be seen that the second definition includes chairs, armchairs and a bed with a sofa, and even a table on which one can also sit. Here our essence combines objects of other entities, other information objects.
Let's go further ... Rules of information interaction.Let's consider, for example, the color indication of a traffic light as a possible system, and select the objects of the system we are considering. The objects of the system are three elements of light: red, yellow, green. If a person does not understand the purpose of a traffic light, then, observing it, he can still see the connections between these objects. Red light comes on after green, yellow after red, green after yellow; sometimes there are traffic lights on which green lights up after simultaneously burning red and yellow, etc. Adding new objects to this system - cars, it will reveal that everyone is going to green, to red - they stop in front of a traffic light, etc.
, — , . ,
One more example. Let's take the entity "Counterparty" - some natural or legal person with whom any activity is carried out; add a new entity to the system - "Deal" . If the entity "Deal" is associated with the "Counterparty" , then we can get 2 new entities - "Buyer" and "Supplier" .
Now, understanding what an information system is, let's analyze what an automated information system is. By analogy with the analysis of the definition above, we can understand that we need to automate information objects and automate information interaction.
It so happened that automation usually means writing a program that most often requires exactly the same actions from a person, and sometimes more, but on a computer. By definition, automation is a kind of shifting of human functions onto machines. So, if a person kept records in a notebook all his life, and you created a program for him, necessarily using "very cool and advanced technology" in which he keeps the same records, but on a computer, then this is not automation.
Let's take an example: our task is to implement accounting of counterparties in the system. In this case, we immediately present a plate with properties and with the appropriate interface. And if I added a counterparty to the system and made a purchase transaction for it, then it was immediately displayed as a supplier. If this process is implemented on paper in tabular forms, then after the transaction is completed, I will have to rewrite the counterparty data from the counterparties table to the suppliers table. That is, thanks to the system, we still automate something.
Therefore, I would define an "automated information system" as a projection of an information system onto technical means
In conclusion, I would like to draw attention to my opinion: in order not to engage in the automation of chaos, start with the design of the system itself, and then in the future the creation of an information system and its automation will give a tangible effect for the user.