RU122796U1 - DECISION MAKING SUPPORT SYSTEM FOR PSYCHOLOGICAL AND PEDAGOGICAL TESTING - Google Patents

Abstract

A decision support system for psychological and pedagogical testing, containing a decision block, a first factual database block, a query input, characterized in that it includes a data input block, an analysis block, a modeling block, a second factual database block, a third factographic block database, data output block, input of the subject, input of prior probabilities, input of tests, input of probabilities of states, input of test parameters, output of the device, while the query input is connected to the first input of the data input block and to the first input of the block; the first factual database and to the first the input of the block is the second factual database and with the first input of the block the third factual database, the input of the subject is the second input of the data input block, the input of a priori probabilities is the second input of the analysis block, the input of tests is the second input of the block the first factual database, the input of state probabilities is the second entrance b loca the second factual database, the input of the test parameters is the second input of the third factographic database, the output of the first factographic database is connected to the third input of the data input block, the output of the second factographic database is connected to the third input of the analysis block, the output of the block is the third factual database data is connected to the second input of the modeling block, the output of the analysis block is connected to the input of the decision block, the output of which is connected to the second input of the data output block, the output of the data input block is connected to the first input of the analysis block, the output of which is connected

Description

The utility model relates to the field of devices for scientific and mathematical modeling and decision support for psychological and pedagogical testing. The primary field of application of the device is modeling for implementing decision support during psychological and pedagogical testing.

A device is proposed that allows simulation to be performed for adaptive testing of a subject’s cognitive abilities and to obtain the required vector Q at the output of the claimed device.

Cognitive ability is a complex and largely integrated concept. Under the cognitive abilities of the test subject, specialists understand [33]: the integral ability to control mental constructions. Further (for technical purposes), by cognitive abilities we mean the ability of the subject to solve the presented test problem in terms of the internal image, anticipating (anticipating) ways to solve this problem from a set of test problems in the process of analyzing the condition (initial data) of the presented test.

The main idea of the work of the device proposed by the authors is as follows.

In fact, the device proposed by the authors allows real-time modeling and decision making to realize the ability to evaluate the knowledge and abilities of the subject (subject) in an adaptively changing order of presentation of tests of a given methodology.

The above effect is achieved through modeling and decision making by using the claimed device.

The test subject, using the claimed device, solves the test tasks by transmitting data about the selected answer option, for example, to a remote server. The test subject has an input tool that allows you to select from a list of answer options. With its help, he types (forms) various sequences of answers, forming the necessary vector R for the claimed device.

The signal received (converted) from the test subject after data entry in the claimed device (i.e., vector R) is compared with previously (pre) calculated model data from the factographic database (FWD) of the claimed device for various groups of subjects represented by a three-dimensional matrix (array) numeric data.

Modeling is carried out in order to determine (identify) other tests of the method possessing the greatest differentiating power for a given subject. For this, the claimed vector M. is formed in the claimed device.

The success of the current task (test), the test is determined as an estimate of the probability of possessing a certain quality (skill), provided that a certain time was spent on solving the problems of the test. In particular, to construct such estimates, the results of integrating the system of ordinary differential Kolmogorov equations with parameters (transition intensities) determined for different groups of subjects are used.

Note that in practice, with a specific implementation of the claimed device, Z is the data obtained, for example, from the FWA and reflecting (showing) the distance — typical results, for example, as the probabilities of staying in the states of the Markov network corresponding to various tasks.

The goal (task) of the expert (device user) by analyzing the Q data received from the data output unit and guided by personal experience, select a test that has the maximum differentiating power for the next presentation. Once this selection is made, the selected next test will be presented to the test subject (subject).

The current test of the subject is completed if the output vector Q is formed as a result of modeling and decision making is transmitted to the output of the claimed device.

Subsequent analysis (processing) of the model data and the result of decision making Q allows us to study in more detail and evaluate the cognitive abilities of the test subject and make a final decision about the test subject.

Let us briefly consider the idea of adaptive testing [17].

It is assumed that the psychological test allows you to divide the subjects into groups according to a specific attribute, for which it was created to evaluate. Hereinafter, the psychological test means a test that evaluates certain abilities, such as the ability to logical thinking, non-verbal intelligence, and so on. The property of such tests is that the answers to the questions of the assignment may be true or false, in contrast to the situation when any answer is correct, as in tests aimed at building a psychological profile.

The dynamics of passing such tests is directly related to the result, for example, the test subject, who gives quick and correct answers to the tasks presented, probably has a high level of assessed ability. In contrast, the subject’s long meditations on assignments indicate his uncertainty in the answers. The process of performing such a test can be represented as a transition from one task to another.

Models for describing the dynamics of these transitions can be represented by oriented graphs in which the vertices (denoted as rectangles) correspond to states (current tasks), and the arcs (indicated as arrows) correspond to transitions.

In the case of models with continuous time, the testing process can be considered as a random walk along the graph with transitions from one state to another according to the directions of the arcs. These transitions are instantaneous and occur at random times.

It is assumed that the following two properties of Poisson event flows are fulfilled for them:

- ordinary (a stream is called ordinary if the probability of occurrence of two or more events during a short time interval is much less than the probability of the appearance of one event during the same time);

- independence of increments (this property means that the number of events falling into two disjoint intervals does not depend on each other).

It can be shown that in the flows under consideration the number of events X falling in any time interval of length r starting at time t is distributed according to Poisson’s law:

where P _{t, τ} (X = m) is the probability of occurrence of m events during the considered interval, a (t, τ) is the average number of events falling into the interval of length τ, starting at time t. Below we will consider only stationary flows (in which a (t, τ) = ητ, η = const). The parameter η is the intensity of the stationary flow. It is equal to the average number of events per unit time. The average length of time between two adjacent events in this case is 1 / η.

The above assumptions about the properties of event flows are common for applied problems, since these flows (or flows close to them in properties) are often encountered in practice due to limit theorems for event flows [40, 41].

Under these assumptions, the dynamics of the time variation of the probabilities of the system staying in different states is described by the Kolmogorov system of ordinary differential equations, in which each state corresponds to the equation

,

where k is the state index; p _k and p _i are the probabilities of being in the kth and ith states; η _kj are the intensities of the flows leaving the state k; η _ik are the intensities of the flows entering the state k. To integrate the specified system, it is necessary to set the initial conditions: p ₀ (0), P ₁ (0), ..., p _n (0). Normalizing condition performed at any given time.

For continuous-time models, the unknown (free) model parameters are the intensities of the event flows. Their values are determined by comparing the observed and predicted histograms describing the distribution of the frequencies of stay in the model states, namely: the values are calculated that provide the best correspondence between the observed and expected frequencies of falling into a certain state of the system at given points in time. The predicted probabilities of being in states are obtained by numerically integrating the systems of Kolmogorov equations (1).

Markov models with continuous time and free parameters, which are identified by observational data, are called Markov networks [39].

To describe how, when passing a given test, the probabilities of being in a given state change over time, Markov networks are used. Figure 10 shows an example of a Markov network. It is assumed that the test consists of a certain number of tasks, each of which can be completed correctly or not. States S _{i, +} (i = 1, ..., m) correspond to the correct execution of the i-th task, and S _{i, -} to its failure or incorrect execution. S ₀ - the initial state in which the test subject is before the test. The intensities of transitions between states a , b, c, d, x, y, w and z are identified separately for each ability level according to the sample of subjects, and x, y, w and z reflect the change in the ability to perform test tasks over time.

The dynamics of the probabilities of being in various states of a specified network is described by the following system of ordinary differential equations:

where P _{S *} (t) is the probability of being in the S _* state at time t. At the start of testing, p _S0 (0) = 1, and the remaining probabilities of being in the states of the Markov network are zero.

To simplify the problem, the flow intensities in some cases can be considered dependent on each other, including the trivial version: x = y = w = z, a = b, c = d. A suitable choice of such dependencies is based on the method of testing statistical hypotheses.

The adaptive testing procedure consists in the sequential presentation of tests from a given set to the subject. Tests are not repeated. The testing procedure is completed upon reaching the specified condition imposed on the probability of the most plausible diagnosis (for example, you can require that this probability exceed 0.7). If the specified condition is not satisfied during the entire testing procedure, then the subject is presented with the entire prepared set of tests.

The psychologist (expert) and the subject work, for example, at different computers between which data exchange is organized. The subject’s monitor is used to present tests, and the psychologist’s monitor is used to display test results and the data needed to select the next test. After the psychologist selects the test, the psychologist appears on the monitor of the subject the corresponding task, which the subject is trying to complete while working at his computer.

After each test is completed, the following are calculated:

- the probabilities of various variants of diagnoses provided that the subject belongs to each of the diagnosed groups and is in the observed state of the network at a control point in time;

- the probabilities of various variants of diagnoses provided that the subject belongs to each of the diagnosed groups and after a predetermined time interval is found in the final state S _{m + of} each of the tests that were not presented at this time.

The strategy of the person making the decision on the sequence of the tests is due to the need to ensure the least differentiation of the test result in the shortest time, represented by the condition imposed on the probability of the most plausible diagnosis. This task does not have a strictly defined solution due to the known uncertainty of the subject’s real actions, including unpredictable time for completing tasks. Therefore, in the absence of any additional arguments, the decision maker should, as a rule, choose the following test as the next one, which leads to the greatest differentiation of the most probable diagnosis from its other options, provided the subject is after a predetermined time interval in the corresponding final state S _{m +} . This approach provides a close to optimal sequence of passing tests and allows various variations of the testing procedure, depending on the experience and additional information available to the psychologist.

The calculation of the probabilities of diagnoses during the next test is performed according to the well-known Bayes formula (2). They are calculated for each of the considered diagnosis options at a control time for the observed state of the network in which the subject finds himself after completing the last task proposed to him:

where C _i is the event associated with the test patient’s i-th variant of diagnosis (i = 1, ..., I), S is the event associated with the network being in the observed state at the control moment in time, P (C _i ) is the a priori probability the presence of the i-th variant of diagnosis, P (S | C _i ) - the probability of being in the observed state of the network at a control point in time with the i-th variant of diagnosis, P (C _i | S) - the probability of the i-th variant of diagnosis the condition of being in the observed state of the network at a control point in time. The probabilities P (S | C _i ) are calculated by numerically integrating the above Kolmogorov equations.

The probabilities of finding at the end of the specified time interval in the final state Sm + of each of the non-presented tests are calculated for all the considered diagnosis options using the Bayes formula [17]: , where C _i is the event associated with the test patient’s i-th diagnosis option (i = 1, ..., I), S _f - is the event associated with being in the state S _{m +} after a specified time interval, P (C _i ) is the a priori probability of the test patient having the i-th diagnosis option, P (S _f | C _i ) is the probability of being in the state S _{m +} after a predetermined time interval in the presence of the i-th diagnosis option, P (C _i | S _f ) - the probability of the i-th variant of the diagnosis, provided that it is in the state S _{m +} after a predetermined time interval. The probabilities P (S _f | C _i ), as in the previous case, are calculated by numerically integrating the Kolmogorov equations. As a priori probabilities P (C _i ), the probabilities P (C _i | S) calculated from the results of the last test performed should be substituted into the formula.

The most likely diagnosis option is the one at which the greatest conditional probability is achieved. calculated from the results of the last completed test. The probability distribution {P (C _i | S)} _{i = 1, ..., I} , allows us to assess the degree of reliability of this estimate [17].

Consider the proposed device in more detail. Further we will also rely on the works [1-15, 16, 17, 18-37].

Known utility model of the Russian Federation No. 114186. "An automated system for monitoring the technical condition and supporting decision-making to improve the safety and reliability of complexes of hydraulic structures of hydroelectric power plants and other facilities" [1], which is a device containing various elements (blocks), as well as: a database, an automated statistical block data processing, source data preparation unit, information input-output means.

This device allows you to solve only one insignificant part of the task, namely, for example, using a block for the preparation of initial data, means for input-output of information to realize a fragment of a block of data input, using a block of automated statistical data processing to implement a fragment of a block of analysis, using a database implement a block fragment first factographic database.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement, the modeling unit, the decision unit, the second factographic database unit, the third factographic database unit, the output unit 8 data i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Another useful model of the Russian Federation is known No. 1055769. "A system for supporting the adoption of management decisions for the formation of the marketing policy of service enterprises" [2], containing various elements (blocks), as well as: a data input unit, a data storage unit, a management decision support unit.

This device allows you to solve only one minor part of the task, namely, for example, using the data input unit to implement the input of some data, using the data storage unit to partially (not fully) store and issue some factual data using the support unit managerial decision making partially (not in full) decision making.

This device does not allow the transfer of the required data to the proposed device, since the following inputs are missing: request input, test input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the implementation of the analysis unit, simulation unit, data output unit 8 in full: decision making unit, first factographic unit database, block the second factual database, block the third factographic database, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow to carry out the necessary conclusion of the results, perform the necessary analysis and modeling in full, make the necessary decision, perform the necessary storage and delivery of factual data, and as a result of this does not allow to model and carry out decision support for psychological and pedagogical testing by forming the output vector Q.

Another useful model of the Russian Federation No. 106397 is known. "Automated information and analytical decision support system for the technical re-equipment of chemically hazardous enterprises" [3], containing various elements (blocks), as well as: data processing unit, data storage unit, visualization unit and output data.

This device allows you to solve only one minor part of the task, namely, for example, using the data processing unit to realize a fragment of the analysis unit, using the data storage unit to partially (not fully) store and issue some factual data using the visualization unit and output data to partially (not fully) output the necessary data.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement the data input unit, simulation unit, decision unit, in full: analysis unit, unit first factographic database block second factual database, block third factual database, block 8 data output, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Another useful utility model of the Russian Federation No. 1133386 is also known. "A decision support system for a distance learning system for specialists in the field of navigation equipment for a GLONASS user" [4], which contains various elements (blocks), as well as a block of subject databases, a block of predictive models.

This device allows you to solve only one insignificant part of the task, namely, for example, using a block of subject databases to partially (not fully) store and issue some factual data, using a block of prognostic models to implement a fragment of a modeling block.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the implementation of the data input unit, analysis unit, decision unit, data output unit 8, in full: simulation unit, block the first factographic database, block the second factual database, block the third factual database, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Known patent of the Russian Federation No. 2294013. "A decision support system to increase the profitability of air transportation" [5], which contains various elements (blocks), as well as an adder, a comparator, an identification block for identifying arrays of data from a group of flights, a server database.

This device allows you to solve only one insignificant part of the task, namely, for example, using an adder, a comparator to implement a fragment of the analysis unit, using the identification unit for the arrays of data of a group of flights, the server database to partially (not fully) store and issue some factual data.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement the data input unit, simulation unit, decision block, data output unit 8, in full: analysis unit, block the first factographic database, block the second factual database, block the third factual database, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known is the patent of the Russian Federation No. 2386569. "Aircraft crew decision support system for the prevention of special situations" [6], containing various elements (blocks), as well as a forecast block, a decision-making calculator on the transition to automatic control, and information display systems.

This device allows you to solve only one insignificant part of the task, namely, for example, using the forecast block to partially implement the modeling block, using a decision calculator to switch to automatic control, implement a fragment of the decision block, using the information display system, partially display the results modeling and decision making by displaying them.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary complete output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the implementation of a data input unit, an analysis unit, a first factographic database unit, a second factual database unit, a third factographic unit database in full: decision block, modeling block, data output block 8, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known is utility model No. 52220. "A decision support system for increasing air transportation revenues" [7], containing various elements (blocks), as well as: a memory block, a code comparison block.

This device allows you to solve only one insignificant part of the task, namely, for example, using a code comparison unit to implement a fragment of an analysis unit, using a memory unit to partially (not fully) store and issue some factual data.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test input, a priori input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement the data input unit, simulation unit, decision block, data output unit 8, in full: analysis unit, block the first factographic database, block the second factual database, block the third factual database, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Utility model No. 57481 is also known. "An automated decision support system in emergency situations" [8], containing various elements (blocks), as well as: a module for receiving user requests, an adder, a module for selecting a reference record address, a data output module.

This device allows you to solve only one insignificant part of the task, namely, for example, using a module for receiving user requests to implement a fragment of a data input unit, using an adder, a module for selecting a reference recording address, implement a fragment of an analysis unit, partially using a data output module ( not in full) output of the necessary data.

This device does not allow the necessary input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test input, a priori input, test input, state probability input, test parameter input.

This device does not allow to fully realize the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement a simulation block, a decision block, a first factographic database block, a second factographic database block, a block the third factographic database, fully necessary, data input unit, analysis unit, data output unit 8, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow full input of the initial data, the necessary output of the results, the full analysis and modeling, the necessary decision to be made, the storage and delivery of factual data, and as a result, it does not allow modeling and supporting the adoption solutions for psychological and pedagogical testing by forming the output vector Q.

Also known patent of the Russian Federation No. 2262131. "A device for modeling the decision-making process" [9], containing different elements (blocks), as well as: two groups of registers, a comparator, a summing block.

This device allows you to solve only one insignificant part of the task, namely, for example, using a group of registers, a comparator, a summing unit, to realize a fragment of an analysis unit.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the implementation of a data input unit, a simulation unit, a decision unit, a first factographic database unit, a second factographic database unit, block third factual database, data output block 8, fully implement the analysis block, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known utility model No. 98387. "A decision support system for the operational dispatching and operational personnel of an automated sorting slide (SPPR KDK SU)" [10], containing various elements (blocks), as well as: a unit for the intellectual analysis of device parameters, a data processing and cleaning unit, a base unit telemetry data, reporting unit.

This device allows you to solve only one insignificant part of the task, namely, for example, using the block for the intellectual analysis of device parameters, the data processing and cleaning block, to realize a fragment of the analysis block, using the telemetry database block to partially (not fully) store and the issuance of some factual data, with the help of the reporting unit, partially (not in full) to output the necessary data.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow to fully implement the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement the data input unit, simulation unit, decision unit, in full: analysis unit, first unit factographic database, block second factographic database, block third factual database, data output unit 8, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known is a utility model No. 23701. "A device for combining criminal cases, determining counterfeit banknotes, securities and documents when solving crimes in forensics" [11], containing various elements (blocks), as well as: a comparison unit, a modeling unit for an automated factographic information retrieval system, an adoption unit solutions, input request.

This device allows you to solve only one insignificant part of the task, namely, for example, using the request input to implement the request input, using the comparison unit to realize a fragment of the analysis unit, using the modeling unit of the automated factographic information retrieval system to partially (not in full ) modeling to support decision making, using the decision block, implement a fragment of the necessary decision block and partially (not in full) IU) necessary decisions.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: test subject, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the data input unit, the first factographic database block, the second factographic database block, the third factographic database block, the block 8 data output, in full: analysis unit, modeling unit, decision making unit, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known is utility model No. 8145 G09B 23/02. "Device for modeling the values of the membership function" [12], containing various elements (blocks), as well as: a memory block, three comparison blocks, a summing block.

This device allows you to solve only one insignificant part of the task, namely, for example, using the comparison blocks and the summing block to realize a fragment of the analysis block, using the memory block to partially (not fully) store and issue some factual data.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement the data input unit, simulation unit, decision unit, in full: analysis unit, unit first factographic database block second factual database, block third factual database, block 8 data output, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known is also a utility model No. 14305 G09B 23/28. "An information system that simulates the algorithm for the occurrence and development of diseases of a non-communicable nature" [13], which represents a device containing various elements (blocks), as well as data processing and storage units, a display unit.

This device allows you to solve only one insignificant part of the task, namely, for example, using data processing and storage units to realize a fragment for the analysis unit, using the display unit to partially display the results of modeling and decision making by displaying them.

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the implementation of a data input unit, a simulation unit, a decision unit, a first factographic database unit, a second factographic database unit, block the third factographic database, in full: analysis block, data output block 8, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Also known is also a utility model No. 20686 G09B 23/02.

“A device for modeling the values of the probability function of the correct answer to the AFIPS forensic inquiry” [14], which contains different elements (blocks), as well as a summing block, first and second probability calculation blocks, a block for outputting simulation results.

This device allows you to solve only one insignificant part of the task, namely, for example, using the summation block, the first and second probability calculation blocks, to realize a fragment of the analysis block, using the block of the simulation results output block, partially display the simulation results and make decisions by displaying them .

This device does not allow the input of all initial data (their transfer to the proposed device) since the following inputs are missing: request input, test subject input, a priori probability input, test input, state probability input, test parameter input.

This device does not allow the necessary full output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow the implementation of a data input unit, a simulation unit, a decision unit, a first factographic database block, a second factual block database, block the third factographic database, in full: analysis unit, data output unit 8, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access the necessary factual databases (FBD) and output the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

Closest to the proposed device is a utility model No. 110638 (prototype) "Navigation device for the orientation of the blind" [15], containing various elements (blocks), as well as: request input, decision block, factographic database block, calculation block, sound indication unit.

This device allows you to solve only one insignificant part of the task, namely, for example, using the request input to implement the request input, using the calculation unit to implement a fragment of the analysis unit, using the sound display unit to partially (not fully) output data (t .e. to partially derive the results of modeling and decision making by indicating them), using one block the factographic database to partially (not fully) store and issue some factographers iCal data by decision blocks to implement desired fragment block the adoption of decisions and partially (not fully) to take a decision.

This device does not allow the input of all the initial data (their transfer to the proposed device) since the following inputs are missing: test subject, a priori probabilities input, tests input, state probabilities input, test parameters input.

This device does not allow the necessary output of all data from the device (there is no necessary output of the data output unit, which is the output of the device) and does not allow to implement the data input unit, modeling unit, in full: decision block, analysis unit, second factographic database unit block third factographic database, block 8 data output, i.e. it does not allow to fully enter data, perform the necessary analysis, perform the necessary modeling, make the necessary decisions, access all the necessary factographic databases (FBD) and make the necessary conclusion of the result of the proposed device.

Thus, this device does not allow the input of input data, the necessary output of the results, to fully perform the necessary analysis and modeling, to make the necessary decision, to store and issue factual data, and as a result of this does not allow to model and perform decision support for psychological and pedagogical testing by forming the output vector Q.

The proposed utility model solves the problem of modeling and decision making based on the results of which the output vector Q is formed as a result of modeling and decision making. And as a result of this, the class of tasks to be solved is expanding.

To achieve the named technical result with a useful model containing a decision block, a first factographic database block, a query input, a data input block, an analysis block, a modeling block, a second factographic database block, a third factographic database block, a data output block, a test input , a priori probability input, test input, state probability input, test parameter input, device output, while

the request input is connected to the first input of the data input unit and to the first input of the unit the first factual database and to the first input of the unit the second factual database and to the first input of the unit the third factual database,

the input of the test is the second input of the data input unit, the input of a priori probabilities is the second input of the analysis unit,

the input of the tests is the second input of the block the first factual database,

the input of state probabilities is the second input of the block; the second factual database,

the input of the test parameters is the second input of the block the third factual database,

the output of the block is the first factographic database connected to the third input of the data input block

the output of the block, the second factographic database is connected to the third input of the analysis block,

block output, the third factographic database is connected to the second input of the modeling block,

the output of the analysis unit is connected to the input of the decision unit, the output of which is connected to the second input of the data output unit,

the output of the data input unit is connected to the first input of the analysis unit, the output of which is connected to the first input of the simulation unit, the output of which is connected to the first input of the data output unit, the output of which is the output of the device.

Signs that distinguish the proposed device from the closest known to it according to the certificate (patent) of the Russian Federation No. 110638 (prototype) [15], characterize the presence of the following blocks, inputs, outputs and connections between them: data input unit, analysis unit, modeling unit, block the second factographic database, block the third factographic database, data output unit, test subject input, a priori probability input, test input, state probability input, test parameter input, device output, while

the request input is connected to the first input of the data input unit and to the first input of the unit the first factual database and to the first input of the unit the second factual database and to the first input of the unit the third factual database,

the test subject is the second input of the data input unit, the a priori probability input is the second input of the unit

analysis

the input of the tests is the second input of the block the first factual database,

the input of state probabilities is the second input of the block; the second factual database,

the input of the test parameters is the second input of the block the third factual database,

the output of the block is the first factographic database connected to the third input of the data input block

the output of the block, the second factographic database is connected to the third input of the analysis block,

block output, the third factographic database is connected to the second input of the modeling block,

the output of the analysis unit is connected to the input of the decision unit, the output of which is connected to the second input of the data output unit,

the output of the data input unit is connected to the first input of the analysis unit, the output of which is connected to the first input of the simulation unit, the output of which is connected to the first input of the data output unit, the output of which is the output of the device, which allows modeling and decision-making to support psychological and pedagogical decision-making testing.

The proposed utility model is illustrated by the drawings shown in figures 1-12.

Figure 1 shows a block diagram of a utility model. The given block diagram reflects the set of necessary blocks for modeling and decision making, and communication between blocks. The block diagram shows:

data entry unit 1, analysis unit 2, modeling unit 3, decision making unit 4, unit 5 first factual database, unit 6 second factual database, unit 7 third factual database, data output unit 8, V - request input, W - test subject input, Y - a priori probability input, A - test input, B - state probability input, G - test parameter input, R - output of data input block 1, E - output of analysis block 2, M - output of modeling block 3, F - the output of decision block 4,

T is the output of block 5 the first factual database,

D - the output of block 6 is the second factographic database,

Z - output block 7 third factual database,

Q is the output of the data output unit 8, which is the output of the device.

Figure 2 ÷ 9 shows a block diagram of the algorithms for each block of the utility model for a particular case (see Example below) for its industrial implementation using programmable computers (for example, personal computers).

Figure 2 shows the block diagram of the algorithm of the program of work of block 1 data input.

Figure 3 shows the block diagram of the algorithm of the program of work of block 2 analysis.

Figure 4 shows the block diagram of the algorithm of the program of work of block 3 simulation.

Figure 5 shows a block diagram of the algorithm of the program of work of decision block 4.

Figure 6 shows the block diagram of the algorithm of the program of work of block 5 of the first factual database.

Figure 7 shows a block diagram of the algorithm of the program of work of block 6 of the second factual database.

On Fig shows a block diagram of the algorithm of the program of work of block 7 of the third factual database.

Figure 9 shows the block diagram of the algorithm of the program of work of block 8 data output.

Figure 10 shows an example of a Markov network.

11 in table A shows the characteristics of the computers (calculators) that tested the device variant, and table B shows the number of bytes sent and received for the server after passing one of the tests.

On Fig shows data on the number of transmitted and received data depending on time.

The proposed utility model works as follows.

Consider her work on the example of solving the modeling problem and making decisions, as a result of which it is possible to carry out decision support for psychological and pedagogical testing by forming the output vector Q.

The inputs of the device (figure 1) receive the source data for its operation: V, W, Y, A, B, G.

It is assumed that the device user (expert, subject) makes his decision V (expresses his desire as a request to the claimed device to receive the output vector Q) either in the form, for example, of a logical zero, or in the form of a logical unit. If the value of V changes from logical zero to a logical one, then this means a request from the device user to receive decision support for psychological and pedagogical testing by forming the output vector Q.

First, at the input of the request, the value of V is a logical zero. It is assumed that after all the initial data has arrived, for example, to the inputs: Y, A, B, G, then at the input of the request, the value of V changes from a logical zero to a logical unit, which means a request from the device user to perform modeling and decision making.

Parameter V through the request input goes to the first input of data input unit 1, to the first input of block 5 the first factual database and to the first input of block 6 the second factual database and to the first input of block 7 the third factual database.

The parameter W through the input of the test goes to the second input of block 1 data input.

The parameter Y through the input of a priori probabilities arrives at the second input of analysis block 2.

Parameter A, through the test input, enters the second input of block 5, the first factual database.

Parameter B, through the input of state probabilities, goes to the second input of block 6, the second factual database.

Parameter G, through the input of test parameters, enters the second input of block 7, the third factual database.

When V is a logical zero, FBDs (first, second, or third) work in write mode, and in the case when V is a logical unit, factual data can be read (from the FBD) - this is reading (unloading) T, D, Z. When V is logical zero FBD works in the recording mode - the factual data is recorded (in the FBD) - this is the recording (loading) A, B, G.

It is assumed that the device user (test subject) makes his decision in the form of the answer number W (i.e., numbers from 1 to m _j ) in problem j. Based on the test results of the test subject, it is required to determine (give recommendations) to which class the subject belongs (there may be K classes in total) and which next test (indicate the number of the M test) should be applied to this subject.

We introduce the following notation (and will use them later):

K is the number of classes of subjects;

L is the number of states of the Markov network;

H is the number of tests;

N is the number of tasks in the test;

m _j is the number of answers in task j;

W is the number (number from 1 to m _j ) of the test subject's answer in task j;

M is the number (number from 1 to H) of the next recommended test for the subject;

R = (r ₁ , r ₂ ) is a vector in which r ₁ is the response time to the last Nth test task; r ₂ is the number of the answer to the last Nth task of the test;

Y = (y ₁ , ..., y _i , ..., y _K ) is a vector for which y _i is the a priori probability of classifying the subject as class i;

E = (e ₁ , ..., e _i , ..., e _K ) is a vector for which e _i is the assessment (according to the test results) of the probability that the subject belongs to class i;

F = (f ₁ , f ₂ ) is a vector for which

f ₁ = max {e ₁ , ..., e _i , ..., e _K } = e _l ;

f ₂ = l;

Q = (q ₁ , q ₂ , q ₃ ) is a vector for which

q ₁ = f ₁ ;

q ₂ = f ₂ ;

q ₃ = M;

T = {T ₁ , ..., T _j , ..., T _N } is a set of N test problems, in which T _j is the condition of problem j with m _j variants of possible answers for the subject;

D = (d _sjt ) is a three-dimensional matrix (array) of numerical data for which

d _sjt is the probability of state s of the Markov network;

s is the state number in the Markov network, where s = 1, 2, ..., L;

j is the class number of the subject, where j = 1, 2, ..., K;

t is the time instant number, where j = 1, 2, ..., K;

Z = {Z ¹ , Z ² } - a set of 2 matrices (arrays) of numerical data, for which

- probabilities for a successful (correct) solution, i = 1, 2, ..., H; j = 1, 2, ..., K;

- probabilities for the case of a not successful (not correct) solution, i = 1, 2, ..., H; j = 1, 2, ..., K;

- the probability for the case of a successful (correct) solution, i = 1, 2, ..., H; j = 1, 2, ..., K;

- the probability for the case of an unsuccessful (not correct) solution, i = 1, 2, ..., H; j = 1, 2, ..., K.

Blocks of the claimed device operate as follows. The data input unit 1 for the input data T, W, and V received by its input forms (if V is a logical unit) the output vector

R = (r ₁ , r ₂ ).

The data input unit 1, based on the received data T from block 5, the first factual database performs the presentation of test questions to the test subject (device user) with the possibility of user selecting answer options. The number of the selected answer option W of the test subject and his response time t _x (i.e., the absolute response time (time since the start of testing) to the last test task, and the answer option) form the vector R = (r ₁ , r ₂ ):

r ₁ = t _x ,

r ₂ = W.

When the answers of the test person are received for all tasks transferred to the data input unit, the vector R is transmitted to the output of the data input unit 1. The vector R is transmitted to the output of the data input unit 1, and then to the first input of the analysis unit 2.

Block 2 analysis of the input data R, D and Y received by its input performs analysis and its result forms the output vector

E = (e ₁ , ..., e _i , ..., e _K ).

Analysis unit 2, based on the obtained factual data D = (d _sjt ) from block 5, the second factual database, the vector Y = (y ₁ , ..., y _i , ..., y _K ), the vector R = (r ₁ , r ₂ ) forms the output vector E = (e ₁ , ..., e _i , ..., e _K ):

,

Where

i∈ {1, ..., K},

Δt is a given constant (Δt ≠ 0 and r ₁ ≥Δt), for example

Δt = 10 seconds

[x] is the integer part of x.

In other words, on the basis of the vector of a priori probabilities (various diagnoses) Y, data on the time and response number in the form of a vector R, results of integration of a general differential Kolmogorov system of general equations (1) [17] in the form of a matrix D, analysis block 2 forms a set of conditional probabilities (2) [17] various diagnosis options in the form of a vector E, as described above.

Vector E is transmitted to the output of analysis unit 2, and then to the first input of modeling block 3 and to the input of decision block 4.

Block 3 simulation of the input data E, and Z performs the simulation of the possible results of other battery tests as follows. The following are calculated:

conditional probabilities M ^{- of} all diagnosis options in case of unsuccessful solution for each test:

, ,

and

conditional probabilities M ^{+ of} all diagnosis options in case of successful solution for each test:

, ,

Then, the sum of the maximum moduli of the component-wise difference is calculated and in the following way:

,

where | x | - the module of the number x, while the final simulation result is the number of the recommended test:

M = h.

The simulation result in the form of the recommended test number M is transmitted to the output of the simulation unit 3 and then to the first input of the data output unit 8.

In decision decision block 4 on the input data E received at its input, a decision is made about the vector F (about which class the subject belongs to) according to the following rule:

f ₁ = max {e ₁ , ..., e _i , ..., e _K } = e _l ;

f ₂ = l;

F = (f ₁ , f ₂ ).

The result of decision F is transmitted to the output of decision block 4 of decision a, then it is transmitted to the second input of block 8 of data output.

In block 5, the first factual database of the input data V and A received at its input is used for initial loading (writing to the FWD) and subsequent transfer (reading from the FWD) of the data as follows.

If V is a logical zero, then this means that block 5, the first factual database is in the recording (loading) mode of factual data. In this case, the current factual data (FD) A is recorded (loaded) A in the FBD of block 5, the first factual database.

If V is a logical unit, then this means that block 5, the first factual database is in the mode of reading (uploading) factual data. In this case, T is sampled (read) from the FBD of block 5, the first factual database.

The result of sampling (reading) T is transmitted to the output of block 5, the first factographic database a, and then to the third input of block 1 data input input.

In block 6, the second factographic database of the input data V and B received at its input is used for initial loading (writing to the FWD) and subsequent transfer (reading from the FWD) of the data as follows.

If V is a logical zero, then this means that block 6 of the second factual database is in the recording (loading) mode of factual data. In this case, the current factographic data (FD) B is recorded (loaded) in the FWD of block 6, the second factual database.

If V is a logical unit, then this means that block 6 of the second factual database is in the mode of reading (uploading) factual data. In this case, a sampling (reading) of D from the FWA of block 6 takes place, the second factual database.

The result of sampling (reading) D is transmitted to the output of block 6, the second factographic database a, and then to the third input of block 2 of the analysis.

In block 7, the third factual database of the input data V and G received at its input is used for initial loading (writing to the FWD) and subsequent transfer (reading from the FWD) of the data as follows.

If V is a logical zero, then this means that block 7 of the third factual database is in the recording (loading) mode of factual data. In this case, the current factographic data (FD) G is recorded (loaded) in the FBD of block 7, the third factual database.

If V is a logical unit, then this means that block 7, the third factual database is in the read (upload) mode of factual data. In this case, the sampling (reading) of Z and FBD of block 7 is the third factual database.

The result of sampling (reading) Z is transmitted to the output of block 7, the third factographic database and, then, to the second input of modeling block 3.

The data output unit 8 according to the input data M and F carries out the output of modeling and decision data for providing them to the user in the form of a vector Q as follows:

q ₁ = f ₁ ;

q ₂ = f ₂ ;

q ₃ = M;

Q = (q ₁ , q ₂ , q ₃ ).

The output of the data output unit 8 is the output of the device.

Thus, the proposed device allows you to successfully perform the necessary modeling, make decisions, and as a result of this allows you to carry out decision support for psychological and pedagogical testing by forming the output vector Q.

The proposed utility model may be industrially applicable (feasible) as follows.

Based on the works [11, 12, 14, 15, 16, 17, 31-37] and on [18-30] we show the possibility of industrial applicability of the utility model proposed by the authors by analogy, as was done previously for other utility models, for example, [11, 12, 14, 15, 35].

In the General case, the blocks of the device can be implemented using a computer (computer) and computer programs.

ROSPATENT made a positive decision to issue a title of protection and the authors successfully obtained certificates for programs and databases of the Russian Federation (RF), for example, [31, 32, 36].

These programs and databases can, with a certain adaptation or modification, be successfully used in the proposed device of the utility model.

An example (industrial application (implementation) of a utility model). Each block of the utility model (Fig. 1) is a standard programmable computer (for example, a personal computer (PC) or a pocket PC (PDA)). Since the proposed device has only 8 pieces of blocks, then 8 pieces of programmable calculators, for example, computers, will be required.

All these blocks (computers) are connected to a standard computer network, for example, on a twisted pair cable using the required number of corresponding network cards and network hubs - HUBs (hubs) [18], with the required number of branches. Another way of connecting blocks (computers) is the well-known, standard way of connecting through serial (COM1-COM4) and (or) parallel ports (LPT1, LPT2) using appropriate standard cables and software, for example, Microsoft [19], Symantec and etc. Each block (computer) is loaded with its own program that implements its function of this block.

Note that it is possible to connect computers (calculators) also using USB ports and the necessary equipment.

Note that it is also possible that these calculators of small sizes (for example, PDAs) are connected to a common computer network (for example, a wireless network) with shared access to some external data storage device (database).

Note that the preparation by the user (subject, expert or administrator) of the files fV№, fV2, fV3, fV4 (each of which contains V, i.e., logical zero (V = 0) or logical unit (V = 1)), means the desire of the user, for example, if the value of V changes from a logical zero to a logical unit, then this means the user’s request for modeling and decision-making (in practice, instead of a file, a variable (flag) in the computer’s memory can be used, which takes either 1 ( logical unit), or value e 0 (logical zero)). Further, we assume that the user (expert or administrator) prepares the files fV1, fV2, fV3, fV4 - containing each V.

Further, we assume that the analysis unit 2 program prepares files fE1, fE2 — containing each vector E.

The device requires the files fV1, fV2, fV3, fV4, fW, fY, fA, fB, fG and fR, fE1, fE2, fM, fF, fT, fD, fZ with data.

The device user (expert), for example, on his personal computer (calculator), which is also connected to a common network, prepares the source files: fV1, fV2, fV3, fV4, fW, fY, fA, fB, fG. Other files: fR, fE1, fE2, fM, fF, fT, fD, fZ and fQ are formed (created) during the operation of the claimed device. All these files are transferred, respectively:

fV1 (containing V) - to the input of the program for block 1 data input,

fV2 (containing V) - to the input of the program of block 5, the first factual database,

fV3 (containing V) - to the input of the program of block 6 the second factual database,

fV4 (containing V) - to the input of the program of block 7 is the third factual database,

fW (containing W) - to the input of the program block 1 data input,

fY (containing Y) - to the input of the program of analysis unit 2,

fA (containing A) - to the input of the program of block 5, the first factual database,

fB (containing B) - to the input of the program block 6 the second factual database,

fG (containing G) - to the input of the program of block 7 is the third factual database,

fR (containing R) - input program block 2 analysis,

fE1 (containing E) - to the input of the program for block 3 modeling, fE2 (containing E) - to the input of the program for block 4 decision,

fT (containing T) - to the input of the program block 1 data input,

fD (containing D) - input program block 2 analysis,

fZ (containing Z) - to the input of the program of block 3 simulation,

fM (containing M) - to the input of the program block 8 data output,

fF (containing F) - to the input of the program block 8 data output,

fQ (containing the output vector Q) is the output of the program of the data output unit 8, the output of which is the output of the device.

It is possible that, for example, files fD, fZ, fY, fA, fB, fG are generated once, for example, by the administrator (expert, device user), and contain default data recommended for permanent use, which can be stored in long-term memory corresponding blocks and do not delete after each reading of data from them. It is possible that, for example, the files fV1, fV2, fV3, fV4 do not have to be frequently deleted by the program using them.

Before starting the operation of the device, not a single block or program is running.

The general operating procedure of the device blocks is as follows:

- if there is a logical zero at the input of the request V:

the program (Fig. 6) of block 5 is launched; the first factual database, which prepares the request for the FWA, (if necessary, recording (clearing) deleting all records from the FWD) recording (loading) the current factual data (FD) A in the FWD;

the program (Fig. 7) of block 6 is launched; the second factographic database, which prepares the request for the FWA, (if necessary, recording (clearing) deleting all records from the FWD) recording (loading) the current factual data (FD) B in the FWD;

the program (Fig. 8) of block 7 is launched; the third factual database, which prepares the request for the FWA, (if necessary, recording (clearing) deleting all records from the FWD) recording (loading) of the current factual data (FD) G in the FWD;

- if there is a logical unit at the input of the request V:

the program (Fig. 6) of block 5 is started (if it had not been started earlier); the first factual database that prepares the request for the FWA, reads (unloads) the required factual data (FD) T, writes this data T to the file fT;

the program (Fig. 7) of block 6 is launched (if not previously launched); the second factographic database, which performs the preparation of the request for the FWA, reads (unloads) the required factual data (FD) D, writes this data D to the file fD;

the program (Fig. 8) of block 7 is launched (Fig. 8) of the third factographic database, which prepares the request for the FWA, reads (unloads) the required factual data (FD) Z, writes this data Z to the file fZ.

If there is a logical unit at the input of the request V, the program (Fig. 2) of the data input unit 1 is launched, then the program (Fig. 3) of the analysis unit 2 is launched, then the program (Fig. 4) of the simulation unit 3 is launched, and then the program (Fig. .5) of decision block 4, then the program (Fig. 6) of block 5 of the first factographic database (Fig. 6) is launched, then the program (of Fig. 7) of block 6 of the second factual database is launched (if not previously run) data, then it starts (if it was not started earlier) programs and (Fig. 8) of block 7 is the third factual database.

The synchronization of the work of programs of all blocks can be carried out, for example, by checking whether each program has a corresponding file (or files) with data and then (if necessary) deleting this file (or files), as well as by checking for a logical zero or logical unit of a special flag (variable).

Before starting the operation of the device, no blocks or programs of blocks are started. The user sets the value of V as a logical zero or as a logical unit. If there is a logical unit (or logical zero) at the input of the V request, the device blocks (programs in them) perform all the necessary actions in accordance with the algorithms for their operation (Figs. 2, 3, 4, 5, 6, 7, 8, 9) .

First, factual databases are prepared for modeling and decision making.

Block 5 is launched first, the first factographic database (Fig. 1), and in it is its program (Fig. 6), with V = 0, and A is the data for the FWA (the factual data A is loaded into the FWA).

The second starts block 6, the second factual database (Fig. 1), and in it is its program (Fig. 7), with V = 0, and B is the data for the FWA (the factual data B is loaded into the FWA).

Block 7 starts the third factual database (Fig. 1), and in it is its program (Fig. 8), with V = 0, and G is the data for the FWA (the factual data G is loaded into the FWA).

Then, the data input unit 1 (Fig. 1) is launched, and in it is its program (Fig. 2), which deletes the file fR if it exists. After that, the presence of the files fV1, fT and the data in them is checked. If these files are not prepared, then checking is performed again, etc. until these files are prepared or the data input unit 1 is turned off. If these files are prepared, then the necessary data is read.

Then, the value of V is checked (if V = 1 (i.e., a logical unit), then this means that the test subject (device user) was asked to model and make a decision).

If V = 0, then the necessary files are checked again, and if V = 1, the following actions are performed.

Checks for the availability (preparedness) of the fW file and the data in it. If this file is not prepared, then verification is performed again, etc. until this file is prepared or the data entry unit 1 is turned off. If this file is prepared, then the necessary data is read (read: W from the fW file).

A vector R- (r ₁ , r ₂ ) is formed as described above. Writing R data to the created fR file.

Next, the files IV1, fW, fT are deleted.

Note that the file fV1 can be stored in the long-term memory of the data input unit 1 and is replaced only if it is necessary to change the program parameters of the data input unit 1.

Then the analysis unit 2 (Fig. 1) is launched, and in it is its program (Fig. 3), which deletes the files fE1, fE2 if they exist. After that, the presence of the files fR, fY, fD and the data in them is checked. If these files are not prepared, then checking is performed again, etc. until these files are prepared or analysis unit 2 is turned off. If these files are prepared, then the necessary data is read. The analysis of the input data and the formation of the vector E = (e ₁ , ..., e _i , ..., e _K ) are performed as described above. Then, data E is written to the created files fE1, fE2.

Next, the files fV1, fW, fT are deleted.

Then, the modeling block 3 (Fig. 1) is launched, and in it is its program (Fig. 4), which deletes the file fM if it exists. After that, the presence of files fE1, fZ and data in them is checked. If these files are not prepared, then the check is performed until the files are prepared or the simulation unit 3 is turned off. If these files are prepared, then the necessary data is read from them.

The necessary modeling is performed as a result of which the value of M is formed (as described above).

After the simulation is performed, the program of the simulation unit 3 saves the simulation result M to the file fM and then deletes the files fE1, fZ.

Note that the simulation result M is transmitted to the input of the program of the data output unit 8.

Then the decision block 4 (Fig. 1) is launched, and in it is its program (Fig. 5), which deletes the file fF if it exists. After that, the presence of the fE2 file and the data in it is checked. If this file is not prepared, then the check is performed until the file is prepared or the decision block 4 is turned off. If this file is prepared, then the necessary data is read from it.

The necessary decision is made as a result of which the vector F is formed (as described above).

After the decision is made, the decision block 4 program saves the decision result F to the file fF and then deletes the file fE2.

Note that the decision result F is transmitted to the input of the program of the data output unit 8.

Then, for block 5, the first factual database (Fig. 1), and in it, for its program (Fig. 6), the user (expert) changes the contents of the V - input of the request (i.e. the file FV2), so V = 1 .

If V is a logical unit, then this means that block 5, the first factual database is in the mode of reading (uploading) factual data.

If the files fV2, fA are prepared and the data in them (and V = 1), then the necessary data is read.

Then there is a sampling (reading) of T from the FWD. Next, the program of block 5, the first factographic database, saves the result T to the file fT and then deletes the files fV2, fA.

Then, for block 6, the second factographic database (Fig. 1), and in it, for its program (Fig. 7), the user (expert) changes the contents of the V - input of the request (i.e. file fV3), so that V = 1 .

If V is a logical unit, then this means that block 6 of the second factual database is in the mode of reading (uploading) factual data.

If the files fV3, fB are prepared and the data in them (and V = 1), then the necessary data is read.

Then there is a selection (reading) D from the FWA. Next, the program of block 6, the second factographic database saves the result D to the file fD and then deletes the files fY3, fB.

Then, for block 7, the third factual database (Fig. 1), and in it, for its program (Fig. 8), the user (expert) changes the contents of the V - input of the request (i.e., file fV4), so V = 1 .

If V is a logical unit, then this means that block 7, the third factual database is in the read (upload) mode of factual data.

If the files fV4, fG are prepared and the data in them (and V = 1), then the necessary data is read.

Then there is a selection (reading) of Z from the FWA. Next, the program of block 7, the third factographic database saves the result Z to the file fZ and then deletes the files fV4, fG.

Then, the data output unit 8 (Fig. 1) is launched, and in it is its program (Fig. 9), which deletes the fQ file if it exists. After that, the presence of fF, fM files and data in them is checked. If these files are not prepared, then the check is performed until the files are prepared or the data output unit 8 is turned off. If these files are prepared, then the necessary data is read from them.

The necessary data output is performed as a result of which the vector Q is formed (as described above).

After that, the program of the data output unit 8 saves the vector Q to the file fQ and then deletes the files fF, fM.

Note that the file fQ containing the vector Q (the results of modeling and decision making) is the end result of the proposed device and is its output (device output).

In order to once again carry out the work of the proposed device (Fig. 1) with other initial data, it is necessary to first prepare the initial (data), for example, files: fV1, fV2, fV3, fV4, fW, fY, fA, fB, fG with data - and after that, then start the first factographic database in block 5 in the device, and its program in it, then start the second factual database in block 6, and its program in it, and then run the block in the device 1 data entry, and in it - its program, then run analysis unit 2 in the device, and in it is his program, i.e. in order as described above.

Note that it is possible to simplify this process of launching these programs in device blocks by a single command using a specially created program (program manager), which automatically monitors the sequence and frequency of program launch for each block of the claimed device.

Thus, the proposed device allows you to successfully perform the necessary modeling, make decisions, and as a result of this allows you to carry out decision support for psychological and pedagogical testing by forming the output vector Q.

Experimental device testing

The authors have developed an experimental model of the dried device to support decision-making for psychological and pedagogical testing by forming the output vector Q.

The device proposed by the authors in practice (in order to verify and confirm the possibility of industrial application (implementation) of the utility model, the claimed device) was implemented as a set of computer programs of a personal computer (calculator) that implements the operation of the algorithms of each block of the claimed device in a graphical programming environment National Instruments Lab VIEW 2010.

The developed software (using, for example, the graphic programming language "G" [37, 38]) is intended for the industrial implementation of the claimed device.

The developed decision support system for psychological and pedagogical testing (DSS DEP) was tested for stability (sustainability) of work. The testing was carried out as follows: the DSS software routine located on the client machine was launched in the standby mode of the service command from the server. The waiting time was limited to 10 minutes (by software this time can be changed according to need). Next, the server part of the DSS software was launched. After sending the command to present the test, the client application analyzed the command, the test was presented and executed (by the test subject), then the necessary data was sent to the server. When transferring packets over the network, no errors were detected - all packets were delivered, the program did not interrupt its execution. The characteristics of the computers (calculators) on which the DSS of the DSS were tested are shown in Fig. 11 in table A.

Network communication between computers was provided by the ASUS WL-520GC router.

When testing the DSS software, the software also used the built-in capabilities of the LabView environment. Elements that provide data transfer via TCP (TCP Read, TCP Write) have the ability to record the number of bytes transmitted and received during program operation. Comparison of these values on the server and client side makes it possible to assert the stability of the DSS software. Figure 11 shows table B, which contains data values that make up the history of states (on the number of bytes sent and received for the server) after passing one of the tests.

The amount of data transmitted over the network directly depends on the number of tasks in the test, as well as on the values of the times for which the task was completed. Within 20 minutes, the DSS SPT was launched, and programmed tests were performed on it. The amount of data (in kilobytes (Kb)) transmitted over this period of time is shown in FIG.

As can be seen from Fig. 12, for 20 minutes of intensive use, the computer (calculator) received 1366.92 Kb of data, 600.35 Kb of data were sent to the client. The difference in the amount of transmitted and received data is related to the amount (amount) of information transmitted: on the server side, this is the only team that identifies the test for presentation to the subject, on the client side, this is the history of states, time intervals for completing tasks and the correctness of tasks.

The fixing of these experimental data was performed on the server computer (computer) of the DSS SPT using the well-known utility (special program) for monitoring NetLimiter traffic.

In general, according to the test results, it can be concluded that the DSS DFT works stably with the required system capabilities of the computers (calculators) on which it is installed, as well as with stable network interaction between computers that are the platform for the testing procedure.

According to the results of the experiments, it was established that the claimed device is operational, all detected errors in the programs were corrected during debugging of the programs, and if necessary, the modification of the programs was successfully completed. The claimed device allows you to successfully perform the necessary modeling, make decisions, and as a result of this allows you to carry out decision support for psychological and pedagogical testing by forming the output vector Q.

It is important to note that experimental studies can be successfully carried out for other conditions (for another possible industrial implementation) of the claimed device (utility model).

The proposed utility model may be industrially applicable (feasible) in another way as follows.

The blocks of the proposed utility model, as in devices (utility models) [11, 12, 14, 15, 35], are implemented using (see [20-30]) analog and digital elements (integrated circuits).

This does not exhaust the whole variety of possibilities for the application (construction) of the proposed utility model.

Used sources

1. Patent (certificate) for a utility model of the Russian Federation No. 114186 G06F 17/00. An automated system for monitoring the technical condition and supporting decision-making to improve the safety and reliability of hydraulic systems. structures of hydroelectric power plants and other facilities, publ. 03/10/2012; Bull. No. 7 (analog).

2. Patent (certificate) for a utility model of the Russian Federation No. 1055769 G06Q 90/00. Support system for making managerial decisions of the formation of the marketing policy of service enterprises, publ. 06/20/2011; Bull. No. 17 (analogue).

3. Patent (certificate) for a utility model of the Russian Federation No. 106397 G01W 1/06. Automated information and analytical decision support system for the technical re-equipment of chemically hazardous enterprises, publ. 07/10/2011; Bull. No. 19 (analogue).

4. Patent (certificate) for a utility model of the Russian Federation No. 113386 G06F 17/00. Decision support system for a distance learning system for specialists in the field of navigation equipment of a GLONASS user, 02/10/2012; Bull. No. 4 (analogue).

5. Patent of the Russian Federation No. 2294013 G06F 17/40. Decision support system for increasing the profitability of air transportation, 02.02.2007; Bull. No. 5 (analog).

6. Patent of the Russian Federation No. 2386569 B64C 13/18. Aircraft crew decision support system for the prevention of special situations, publ. 04/20/2010; Bull. No. 11 (analogue).

7. Patent (certificate) for a utility model of the Russian Federation No. 522220 G06F 17/30. Decision support system for increasing air transportation revenues, publ. 03/10/2006; Bull. No. 7 (analog).

8. Patent (certificate) for a utility model of the Russian Federation No. 57481 G06F 12/00, G06F 15/00. Automated decision support system in emergency situations, publ. 10/10/2006; Bull. No. 28 (analogue).

9. Patent of the Russian Federation No. 2262131 G06F 17/00, G06N 7/06. Device for modeling decision-making, publ. 10/10/2005; Bull. No. 28 (analogue).

10. Federation No. 98387 B61L 17/00, G06F 17/00. Decision support system for the operational dispatching and operating personnel of an automated sorting slide (DSS KDK SU), publ. 10/20/2010; Bull. No. 29 (analogue).

11. Patent (certificate) for a utility model of the Russian Federation No. 23701 G07D 7/00. A device for combining criminal cases, determining counterfeit banknotes, securities and documents when solving crimes in forensics, publ. 06/27/2002; Publ. Bull. No. 18 (analog).

12. Patent (certificate) for a utility model of the Russian Federation No. 8145 G09B 23/02. Device for modeling values of membership functions, publ. 10.16.1998 (analogue).

13. Patent (certificate) for a utility model of the Russian Federation No. 14305 G09B 23/28. Information system that simulates an algorithm for the occurrence and development of diseases of non-infectious nature, publ. 07/10/2000 (analogue).

14. Patent (certificate) for a utility model of the Russian Federation No. 20686 G09B 23/02. A device for modeling the values of the probability function of the correct answer to a query AFIPS forensic appointment, publ. 11/20/2001; Bull. No. 32 (analogue).

15. Kuravsky L.S., Kulik S.D., Marmalyuk P.A., Yuryev G.A. Utility Model Patent No. 110638, Russian Federation (RU), cl. IPC ⁸ A61F 9/08. Navigation device for the orientation of the blind / MGPPU (Russia). - Application No. 2011125847; Application 06/24/2011; Priority from 06.24.2011; Bull. No. 33 (prototype).

16. Kuravsky L.S., Yuryev G.A. The use of Markov models in the processing of test results // Questions of Psychology, 2011. - No. 2. - p. 98-107.

17. Kuravsky L.S., Yuryev G.A. Adaptive testing as a Markov process: models and their identification // Neurocomputers: development and application, No. 2, 2011, p.21-29.

18. Ethernet user's manual for CN8800TPC // CNet total network solutions: CNet, 1993-20p.

19. Microsoft solutions. Issue 1-5: Microsoft, 1996.

20. Lysikov B.G. Arithmetic and logical foundations of digital machines. - Mn .: Higher School, 1980.

21. Savelyev A.Ya. Arithmetic and logical foundations of digital machines. - M.: Higher School, 1980.

22. Soloviev G.N. Arithmetic computer devices. - M: Energy, 1978.

23. Evreinov E.V., Butylsky Yu.T., Mamzelev I.A. et al. Digital and Computer Engineering: Textbook for High Schools. - M: Radio and communications, 1991.

24. Shilo V.L. Popular digital circuits. - M.: Radio and Communications, 1987. - 352 p.

25. Bureev L.N., Dudko A.L., Zakharov V.N. The simplest microcomputer. - M .: Energoatomizdat, 1989.-216 p.

26. Tarabrin B.V., Lunin L.F., Smirnov Yu.N. et al. Integrated circuits Reference. - M .: Radio and communications, 1983.

27. Alginin B.E. A circle of electronic automation. - M .: Education, 1990.

28. Hawkins G. Digital Electronics for Beginners. - M.: Mir, 1986.

29. Mukhitdinov M., Musaev E.S. Light emitting diodes and their application. - M .: Radio and communications, 1988

30. Yakubovsky S.V., Barkanov N.A., Nisselson L.I. and others. Analog and digital integrated circuits. - M .: Radio and communications, 1985.

31. Certificate for the program of the Russian Federation No. 2007714119 "Program of decision making and data presentation for a decision maker in the financial market" (PInterface) / S. D. Kulik et al. (Russia). - Application No. 20077613119; Application 07/31/2007; Zaregistr. 10/26/2007; Publ. Bull. No. 4 (61). - Part 2. - S.303.

32. Certificate for the database №2001620076, Russian Federation, "Factographic database AFIPS" Abris "" (FBD-ABR $) / S.D. Kulik (Russia) .- Application No.2001620050; Application 03/29/2001; Zaregistr. 05/25/2001. Bull. No. 3 (36). - S. 348-349.

33. Palmer J., Palmer L. Evolutionary Psychology. Secrets of the behavior of Homo sapiens. - St. Petersburg: Prime-Euroznak, 2003 .-- 384 p.

34. Kuravsky L.S., Margolis A.A., Yuryev G.A., Marmalyuk P.A. The concept of a decision support system for psychological testing problems // Abstracts. X All-Russian Scientific Conference "Neurocomputers and their Application" NKP-2012. Moscow, March 20, 2012 - M: MGPPU, 2012 .-- P.49.

35. Kuravsky L.S., Kulik S.D., Marmalyuk P.A., Yuryev G.A. Utility Model Patent No. 110269, Russian Federation (RU), cl. IPC ⁸ A61F 9/08. A device for orienting the blind. - Application No. 20111120734/14; Application 05/24/2011; Zaregistr. 11/20/2011; Priority from 05.24.2011. Publ. Bull. Number 32.

36. Kuravsky L.S., Yuryev G.A. Certificate for the program of the Russian Federation No. 2009613028 "Program for the recognition and dubbing of texts for people with visual impairment v.1.0" (El-Reader) / L.S. Kuravsky, G.A. Yuryev (Russia). - Application No. 20099611751; Application 04/21/2009; Zaregistr. 06/10/2009. - (ROSPATENT).

37. Dumin P.N. Software implementation of the prototype decision support system for the tasks of psychological testing // Abstracts. X All-Russian Scientific Conference "Neurocomputers and their Application" NKP-2012. Moscow, March 20, 2012 - M: MGPPU, 2012 .-- S.50-51.

38. Peych L.I., Tochilin D.A., Pollak B.P. LabView for beginners and professionals. - M .: Hot Line-Telecom, 2004.

39. Kuravsky L.S. and Baranov S.N. The concept of multifactor Markov networks and its application to forecasting and diagnostics of technical systems // Proc. Condition Monitoring 2005, Cambridge, United Kingdom, July 2005 .-- pp. 111-117.

40. Ovcharov L.A. Applied problems of queuing theory. - M.: Mechanical Engineering, 1969.

41. Saati T.L. Elements of queuing theory and its applications. - M .: LIBROCOM, 2010.

Claims (1)

A decision support system for psychological and pedagogical testing, comprising a decision block, a first factual database block, a query input, characterized in that a data input block, an analysis block, a modeling block, a second factual database block, a third factual block are entered into it database, data output unit, test subject input, a priori probability input, test input, state probability input, test parameter input, device output, while the request input is connected to the first input of the data input unit and to the first input of the unit the first factual database and to the first input of the unit the second factual database and to the first input of the unit the third factual database, the test subject is the second input of the data input unit, the input of a priori probabilities is the second input of the analysis unit, the input of the tests is the second input of the block the first factual database, the input of state probabilities is the second input of the block; the second factual database, the input of the test parameters is the second input of the block the third factual database, the output of the block is the first factographic database connected to the third input of the data input block the output of the block, the second factographic database is connected to the third input of the analysis block, block output, the third factographic database is connected to the second input of the modeling block, the output of the analysis unit is connected to the input of the decision unit, the output of which is connected to the second input of the data output unit, the output of the data input unit is connected to the first input of the analysis unit, the output of which is connected to the first input of the simulation unit, the output of which is connected to the first input of the data output unit, the output of which is the output of the device.