Annotation of the results obtained in 2023
This year, work to achieve the goals of the grant and solve the problems that are planned for the first year has been mainly concentrated around the construction of various multi-agent models reflecting various crisis situations and responses to them within the organization. The first model developed examines the process of pricing and introducing a company's new product to the market. The structural-dynamic model consists of agents, environment, interactions and rules of behavior. In our model, there are two types of agents: firms and clients. Each consumer has a different weight for each type of preference, so consumers are considered heterogeneous. On the part of the company, the following characteristics can be determined: the price of the product (or service), the quality of the services provided, the size of the discount for various categories of clients (in the case of creating a model, we will consider one averaged parameter, so as not to complicate the model), the loyalty of the company’s clients. Agent firms can behave either homogeneously or heterogeneously when making pricing decisions. A firm's pricing decisions involve price promotions determined by the frequency and level of price reductions. The pricing method applies a cost-plus strategy, while price incentives reduce the profit target. The promotion strategy is divided into three levels: high-superficial, moderate and low-deep. According to the idea that we are comparing 3 possible pricing strategies in pairs, we are dealing with 6 possible cases: • Homogeneous (homogeneous) cases: o all companies use the “high-shallow” strategy; o all companies adhere to a moderate strategy; o all companies follow a low-deep strategy; • Cases of heterogeneity: o companies use moderate/high-superficial strategies to some extent; o companies to some extent apply a moderate/low-depth strategy; o companies apply “high-shallow”/ “low-deep” strategies to some extent. The results show that when both retailers compete equally, the resulting results are consistent with Hotelling's law. The replicated data results from the simulation results of 100 repetitions for homogeneous cases indicate (as measured by Welch's t-test at the 95% confidence level) that the difference in average profits between companies is not significant. A similar emerging pattern applies to overall sales. The reason for this small difference in profits and sales between retailers may be due to the identical price promotion strategy used by the retailers. At the same time, Welch's t-test for cases of heterogeneity showed a significant difference in the average profit received by companies. Comparing the moderate strategy with the high strategy and the shallow strategy, we see that at the beginning of the simulation, each strategy alternately dominated profits and sales during the simulation. By comparing the high shallow and low depth strategies, we see that the high shallow depth strategy results in higher profits and sales than the low depth strategy. Finally, for the moderate and low depth strategies, no strategy dominated in terms of profits and sales during the simulation. The second multi-agent model included an analysis of the organization's life cycle in the context of the possible occurrence of various crisis situations during development. In accordance with Adizes, in the developed model we identified 10 stages of the organization’s life cycle. The interaction between agents is illustrated as follows. The model develops cyclically, at each iteration of the cycle the company performs one of the business operations, and the probability of success of the operation is checked. Possible business transactions include the following: • “Improve a business idea”, • “Carry out an advertising campaign”, • “Optimize processes”, • “Carry out restructuring”, • “Improve the product.” First of all, random scenarios of company development were analyzed, without explicit intervention from users. In this case, from the modeling results, 100 repetitions for homogeneous cases indicate (according to Welch’s t-test with a confidence level of 95%) that in the vast majority of cases (~91%) the company does not go through the development stage, but stops there. And only 9% of companies are implemented to the end and reach the death phase after a long (more than 25) years of existence. In the case of obvious interference in the work of the company (which most often happens in practice), 2 directions are clearly distinguished: if at the initial moments of the company’s life the intervention included expensive and time-consuming actions, then the company achieved death in infancy in 73% of scenarios based on the results of more than 10,000 starts. If the company’s development occurs in a balanced manner: in the first 4 stages, “soft” development scenarios are applied (adjustments to the pricing policy, changes in the market niche, etc.), then the company successfully reaches the peak of its development and further its development is possible with elements of restructuring for the transition to subsequent stages of the life cycle. Finally, based on the results of the analysis of the developed models, they were generalized with the construction of a generalized conceptual model to describe the structure and dynamics of IT-intensive business processes in the global digital environment. Since we are considering different types of organizations in the context of crisis situations, in the conceptual model we have summarized the existing experience in solving crisis situations of various types of organizations in the global digital environment. The final conceptual model includes 458 entities of various classes, interconnected by 535 relationships. At the first stage of ontology development, the following entity-organizations were identified within the OWL ontology - a total of 15 entities: • educational organization; • commercial organization; • government organization (including organizations at various levels); • international organization; • organizations that do not engage in income-generating activities, etc. These types of organizations are high-level entities that describe the most common types of organizations in terms of their form and legal affiliation. In addition to this, 22 entities were identified that describe participants in various processes of the organization, in terms of their degree of involvement in the organization’s business processes: • clients of the organization (including a situation where the client may be a minor and his legal representative-parent); • employees of the organization; • participants in company projects (external to it); • students; • visitors (tourists, etc.), etc. Finally, various forms of legal interaction between the company’s clients and the organization itself were highlighted, since they significantly influence possible solutions in crisis situations (from the amount of payments in the case of negative scenarios for resolving the crisis to possible consequences at the level of reputational costs). After constructing a generalized conceptual model in the form of a formal ontology, expressions in a formal language were developed to specify the process and results of the composition of single-aspect multi-agent models. As a result, the process and results of the composition of multi-agent models are presented in the form of thirty formal descriptions.

 

Publications

---