Modeling of objects, functions and operations. Merological relations between objects of this type

    In past articles, I looked at modeling objects, types, and attributes.

    Concepts: set, type, attribute
    How to confuse analytics. Part one
    How to confuse analytics. Part two: what is domain modeling?
    How to confuse analytics. Part three. Verbs and Numerals

    In brief, we touched on the life cycle of an object in terms of its transformation and the transformation of our ideas about it.

    How to confuse analytics - 4. Probability and accuracy

    Next, I began to consider modeling operations, functions and objects from a single point of view.

    How to confuse analytics - 5. Conceptual apparatus

    All this goes beyond the scope of modeling when creating information systems, but to solve the problems of integrating various information systems through the creation of an adapter between them it is extremely necessary.

    To create an adapter, we need to learn how to model the same thing in different ways: as an object and as an action. For philosophical thought, this is not new, because objects do not exist outside of time, and actions cannot be performed without objects. In fact, we have to look at the world as Buddhists look at it: the object and the action are one and the same. The need for such a worldview stems from the need to combine different points of view on the same thing. in this article, I will consider possible representations of reality and the meteorological relations (part-whole relations) between them.

    Watch Description Experiment


    Let there be a clock. We bring three people and ask them to describe what they saw.

    • The first will say that it is an object called a watch.
    • The second will say that it is a function called a demonstration of the current time.
    • The third will say that this is a set of operations under the general name "turn the minute hand by one division."

    All three will describe one four-dimensional space-time object, but they will do this in different ways.

    • In the first case, we get a model of the object,
    • in the second, a model of the function,
    • in the third, an operation model.

    If we want to create an adapter that will connect three different information systems, in each of which the same will be modeled as an object, operation or function, then we need to learn how to change our point of view, and how to see the object in the same amount , and function, and operation. To do this, we must accept the idea that an object, an operation, and a function are different descriptions of the same spatio-temporal volume.

    An object, operation or function is present in the human mind, but does not exist in reality. Not in the sense of solipsism, as the reader might think (the real world exists), but in the sense of interpreting this world, because the object, operation and function are different interpretations of the same space-time volume.

    Modeling reality, we can easily cope with the representation of reality in the form of objects, however, we are almost unable to work with representations of reality in the form of functions and operations. And existing standards do not help us figure this out. In this article, I will examine the relationship between representing reality as an object, function, and operation.

    Modeling begins with a description of the four-dimensional space-time volume, the interpretation of which the analyst then engages in. For this, a model of the boundaries of this volume is constructed - a location in space and in time. For example, the attributes “start date” and “end date” are used to describe time boundaries. After describing the volume, the analyst interprets this volume.

    Representation of reality as an object


    When we interpret the volume of space-time in the form of an object, we focus on statics. This can be form statics, construction statics, or property statics. For example, a chair has a static shape and structure, a river has banks and a water surface, a clock has a structure (the form changes), a chess game has players and a table.

    To describe the construction of an object, a model is constructed in the form of a set of interconnected objects - parts of a simulated object. The volume treated as an object is divided into parts, each of which is treated as an object.

    Representation of reality as a function


    When we interpret the four-dimensional volume of space-time as a function, we focus on the dynamics, which has the property of repeatability. For example, the function of demonstrating time is dynamic, but the events that accompany this dynamics belong to one class - the rotation of the arrow by a given angle (static). That is, the invariant of the function is the class of observable events.

    You can often hear people say that the function of the watch is to show time. This thesis separates the object and the function. In fact, both an object and a function are different interpretations of the same volume. Therefore, from the point of view of modeling, it is right to say: a volume interpreted as a clock can also be interpreted as a function of demonstrating the current time that occurs in a given section of space. Such a thesis leads to the construction of a rather voluminous model, which is why analysts often reduce it to two objects - a function (time demonstration), an object (clock) and the connection between them. This reduces the amount of modeling, but prevents you from thinking correctly. In my interpretation, the model is symmetrical with respect to any of the accounting objects - be it an object, a function, or anything else. The symmetry of the model allows you to build an adapter.

    Functions are modeled using IDEF0 notation. The same notation allows you to model the constructive division of the function into parts - functional structures. This is when a function is divided into parts, each of which is interpreted as a function. Often this model is erroneously called a process model.

    In the same notation, you can see the rudiments of modeling the construction of functions in the form of objects. These are the so-called “arrows from below”. But in this place the notation is incomplete, because it is not clear what these arrows mean - whether the temporal parts of the participants, the interpretation of these parts (roles), or the participants themselves. Therefore, we can say that there are rudiments of modeling, but the notation is incomplete. I will talk about this later in more detail.

    Representation of reality as an operation


    When we interpret the four-dimensional volume of space-time in the form of an operation, we focus on the dynamics in which, in the general case, there are no invariants with respect to time, but there is an invariant with respect to the space in which the operation takes place. That is, operations occur in a certain volume, in which, from the point of view of the subject, causal relationships must be observed.

    Dynamics in the operation is associated with the emergence and destruction of four-dimensional volumes. It can be volumes interpreted as the state of objects, objects, etc. Therefore, the operation model is a set of start dates and completion dates of some four-dimensional volumes.

    The operation must be distinguished from the design of the operation. Usually, an operation model is understood as a specific model of its construction. The encountered constructs of the operation are divided into two types. The first type of construction is a description of its participants, and the second is a construction in the form of sub-operations connected by temporal connections (process). So often, an operation model is understood to mean an enumeration of its participants, that is, its design in the form of participants. The reason for this kind of error was the purpose of modeling. Everyone is interested in the cause-effect relationships that led to the operation, or an explanation of the changes. Causation can be explained by referring to either the obvious composition of the roles or the obvious sequence of sub-operations. This is done in two steps.

    • The first step is to simulate the temporal parts of the operation. If we divide the operation in time, then these will be parts separated in time, which in the future we will treat as operations. If we divide the operation in space, then these will be the parts that exist simultaneously throughout the operation. These parts will be further treated as roles.
    • At the second step, it is necessary, appealing to the subject's empirical experience, to explain to him the changes that occurred in the operation.

    First, consider the design model of the operation in the form of participating roles. To explain cause-effect relationships in this way, it is necessary to indicate the obvious participants who will explain what is happening. For example, if we want to explain why the apple fell to Earth, then the obvious participants in the fall operation will be: the temporal part of the apple, which plays the role of body 1, the temporal part of the Earth, which plays the role of body 2, and the gravitational interaction between the bodies. We referred to the law of universal gravitation and, thus, explained the fall of an object on Earth. To simulate this kind of explanation, it is necessary at the first step to list volumes that will be further treated as temporal parts of objects, for which we need to connect these volumes with those volumes that are interpreted as Earth and an apple. In the second step, assign to them the volume of the role of the participants: the role of body 1 and body 2 in the law of universal gravitation. Then refer to the law and get an explanation. In this case, the model of cause and effect relationships is a model of roles, the model of participants is a model of performers of these roles. Very often you can meet a situation where the model of participants and the model of roles are confused and dumped all in one pile.

    Consider the model of the construction of the operation in the form of a sequence of sub-operations. To explain causation in this way, it is necessary to list the obvious sub-operations that will explain what is happening. For example, to explain the reason how the application from the client was processed, it is necessary to consider the sequence of operations from receiving the application to its archiving. Looking at such a sequence, one can easily understand why all parties interested in this operation were satisfied. To do this, build a sequence of sub-operations and relationships between them. And here again there are two levels - at the first level, temporal volumes and their position in space-time, at the second - the interpretation of these volumes in the form of operations and an explanation of cause-effect relationships between them.

    So far, I have not seen standards in which these levels would be clearly separated. Now the modeling of volumes and their interpretation are piled up in one heap. For example, it is often said that a process should have a result. But the result is the second level of the model. What then is the first level in which there is a sequence, but there is no result, no explanation of cause and effect relationships? Since analysts skip this level of modeling, they don’t need to come up with a name for these kinds of objects. But, building full-fledged models, we are forced to look for these names. Or I recently heard that the role of a business analyst is to help the business. But this is again the second level of the model, at the first level of which there is only modeling activity, but there are no goals and cause-effect relationships. And what then to call the first level, is not clear.

    The mixing of these levels is also facilitated by language and the mythical consciousness that we inherit from our ancestors. When I read that a horse is riding, I can interpret this thesis in three different ways. I can imagine a moving object in the shape of a horse, making specific movements, I can imagine a mechanical system capable of producing such movements, and I can consider the mental functions of the horse. All three modes of presentation have one verb - to jump and no hint of differentiation. That is why analysts in the same model often confuse all three levels: facts, causes and effects, as well as the intentions of the subjects. Unfortunately, there is no way in the language to separate these three different models.

    Different subjects can converge in that they see the same participants in the operation, but diverge in the interpretation of their roles. For example, in war, the warring parties often strongly disagree about their own and other people's role in hostilities. It may be different - subjects recognize the same composition of roles, but name different participants. For example, two masses participate in gravitational interaction, but the applicants for these masses can be different, as is often the case in astronomy.

    Based on this, if we want to build an adapter, we must be able to stitch different points of view on reality, whether it be physical (volumes) or mental (cause-effect relationships). How to build such models, I told earlier, is a separate and very big topic, at the first stage of which the construction of independent physical and mental models is considered.

    Meriological relations between objects of different types


    We saw that the same volume can be represented in the form of objects of different types, we saw that the object can be represented in the form of a construction of objects, a function in the form of a construction consisting of functions or objects, an operation in the form of a construction, consisting of objects, or operations. One may wonder: is it possible to construct an object structure from functions and operations? Is it possible to construct a construction of a function from operations, or a construction of an operation from functions? These are quite specific models, but they also have a place to be.

    • An object - an enterprise can be represented in the form of a design, the elements of which are functions: “equipment production” and “equipment sale”.
    • An object - a clock can be represented in the form of a design consisting of operations “turn the clock hands” (there are a lot of them, but in principle, this is possible).
    • Function - a demonstration of the current time by analogy with the clock can be represented in the form of a design consisting of operations "turn the clock".
    • Representing an operation as a set of functions is also possible. The operation “drive the stake into the ground” can be represented in the form of a design consisting of the functions “keep the stake in the set position” and “beat the sledgehammer at the end of the stake”.

    Thus, not only can any volume be represented as objects of different types, but objects of different types can also be represented as structures consisting of objects of a different type than the modeled object, and there are no restrictions on the types of objects.

    On this I completed the story about the types of objects and the relationships between them. Now you yourself can experiment and find constructions where you previously saw semantic connections. You can also separate the mereological (objective) relationships between objects and imaginary cause-effect relationships, or the intent of the subject.

    How modeling looks in reality


    So far, it seems that everything is logical and simple. However, let's add a little complexity. Recall that the future is modeled using the space of probable outcomes, and modeling of the past is limited by the accuracy of our knowledge. Thus, when we talk about the start date of the existence of a certain object, for example, a photon rocket, we say that this date lies in the future somewhere after 2100. When the year 2050 comes, we will clarify this date and say that the launch date of such a missile is somewhere after the year 2150. Thus, the value of the “start date” attribute will have a spread and this spread will change over time. Therefore, the value of the “start date” attribute is not just a value, but rather a range of values ​​and, since this area changes over time, it should be tied to the date of relevance of this area.

    In systems engineering, it is believed that the life cycle of an object begins from the moment of its design and ends with the moment of its disposal. It is impossible to argue about the time of disposal, but I would argue with the moment of the start of design. During the design process, the planned date of the beginning of the existence of the object is discussed. And this date is not equal to the start date of the design of the object. The planned date of the beginning of the existence of the object is being specified along with the refinement of the design documentation. If we accept that the date of the beginning of the existence of the object is the date of the beginning of its design, then we will get an analogue of the barber shaver paradox, that is, nonsense. The project indicates the date of creation of the facility and its disposal. If in the project it is said that the date of creation of the object is the date of the beginning of its design, then a collision will result.

    Also popular now: