Google Trends is a popular computational linguistics tool designed to determine and analyze the range of public interests, taking into account the linguistic features of the formation of Internet queries. In this paper, it is proposed to use the Google Trends service to determine the risks of cooperative ties between industrial enterprises of the Kharkiv and Zaporizhia regions, which are already leaders in industry clustering in Ukraine. The information base of the study was the groups of enterprises, which were determined by the analysis of the hierarchy of cooperation relations. An analysis of the names of enterprises in the group using Google Trends made it possible to determine the demand for their products by region, which made it possible to assess the state of their promising cooperation. In this aspect, a list of clustering risks is defined, a relative assessment of indicators according to the linguistic features of the formation of scales including. The work carried out an experiment with Kharkіv enterprises, which made it possible to make a decision on cooperation with certain institutions, taking into account the processing of qualitative indicators by using linguistic information. Thus, the scientific novelty of the paper is the technology of processing quality information with widespread use of methods of using linguistic information, which is important for the formation of prerequisites for decision-making in economics was developed by the authors.
Computational linguistics is a relatively new field of knowledge in linguistics, designed to create automated systems for processing textual information for the purpose of its recognition, translation, etc. One of the tasks is to build knowledge management systems, which involves the implications of expert systems to achieve the result of solving an individual problem of natural language processing through mathematical models and artificial intelligence tools. The primary mission of computational linguistics is due to the need to create and improve the linguistic support of information processing systems. The results of the analysis of requests and messages underlying the communication of users with intelligent automated information systems (AIS) can be used not only to achieve the goals of linguistics, but also to process economic databases. AIS solves issues in the economy that require the translation or analysis of language terms, processing the level of interest in these definitions according to certain criteria for belonging to a region, for example, key words and annotations formed using Google, Lingvo and other automatic translators are essential structural elements of an economic article. Computational linguistics is a means of expanding the tools for analyzing economic processes that are dynamically changing in the modern unstable environment of economic development in the country. In the context of the economic crisis caused by epidemiological and political world phenomena, the issue of combating economic risks that have led to a deterioration in the activities of enterprises is becoming increasingly important. Linguistics has the ability to solve the problem of analyzing the real economic situation by determining the current social directions of the country's development. To confirm the priority of this issue in society, a search has begun for ways to overcome risks for enterprises and to analyze the prevailing opinion on this topic in the aspect of the country's regions using Google Trends.
The presented topic is not new for modern scientists, moreover, the number of interested people is constantly growing. So, in 2020, O. Kliuiev [2] with a team of authors developed a successful startup investment model, using in his research the results of interest in the terms “innovation”, “business incubator” in various speech variations. In turn, in 2021, N. Vnukova and S. Achkasova [3] came to the conclusion about the seasonality of interest in financial risk by studying the level of interest in the definition of "forensic" and "risk control" in Google Trends, taking into account discrepancies in time and in dynamics. Also, the results of a study by V. Shmatkov and S. Hlibko [4] are interesting regarding the difference between the concepts of "copyright" and "intellectual property" in the framework of the formation of digital user competence regarding the mechanism for indemnification and the use of results for risk assessment. V. Vysotska [5] dealt with web analytics, including those generated by means of Google Trends, considered e-commerce issues based on the search for relevant information through keywords on the topic based on the feedback activity of the audience. That is, the practice of using Google Trends tools to solve the problems of economic science exists and is widely represented in the studies of modern scientists.
Publication activity on enterprise risk issues is substantiated by the real state of economic development, which is reflected in the publications of young authors. So, M. Klimchuk [6] developed an adaptive model for countering the risks of business processes of alternative energy enterprises and formed a matrix for the distribution of risk zones for it. Y. Shvets [7] analyzed the methods for assessing the risk of industrial enterprises in order to determine the feasibility of their use in the activities of institutions. The topic of enterprise risk assessment is also relevant in foreign studies. For example, A. Hayes [8] considers all the features of risk assessment of modern American enterprises and ways to overcome it. In addition, P. Avila [9] draws attention to the need to assess the risk of collaborative enterprises in order to improve existing models in Portugal. Thus, modern domestic and foreign studies of determining the risk of enterprises are focused on the process of its assessment and methods of prevention. Meanwhile, there are almost no proposals on how to overcome the risks of enterprises, taking into account the means of computational linguistics, so it is proposed to study this issue in more detail using Google tools on the example of Ukrainian enterprises.
In times of unstable development of the financial sector in Ukraine, filled with risks for specific enterprises, a collaborative approach to industrial management that meets the principles of the state policy of the country and the European directions for the development of innovations in its economy is becoming increasingly important. Many industry experts note cooperation as a possible means of overcoming crisis phenomena in the activities of organizations [10]. Cooperation is the association of enterprises to achieve a common goal based on beneficial interaction and consensus. In Ukraine, this idea is actively promoted, which is reflected in the concept of creating clusters by the industrial community of Kharkiv and Zaporizhzha in 2020 [11, 12]. The main mission of these clusters was to establish a partnership between the members of the association in order to strengthen internal interaction with each other and their entry into the international market. In 2021, by order of the German society GIZ (Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH), which supports cluster cooperation in Ukraine, industry experts conducted a study of enterprises in the two regions represented on the possibility of strengthening ties and expanding the circle of cluster members. The authors propose an expert method for analyzing hierarchies [13] as a key research method, based on the structuring of relationships between subjects and their priorities for cooperation. In the constructed hierarchies, the first level symbolizes the high state of potential interaction between enterprises. The opinion of experts was collected through an online Google Form survey of representatives of industry enterprises using the Kharkiv region as an example. The results of the study (fragmentary) are shown in Figure 1. 4 levels are occupied by those enterprises that have prospects for clustering. Enterprise No. 28 “Kommunar Corporation” has all chance to become an active member of the EAM cluster (Engineering, Automation, Mechanical Engineering) through a system of interaction with organizations that already have the opportunity to use the prospects of association through cooperation with other enterprises. The constructed structure makes it possible to determine the directions for expanding the cluster by attracting more promising organizations that are ready for internal cooperation. The depicted hierarchy is difficult to perceive due to the chaotic nature of the directions of interaction, but at the same time it is informative for the logistics system of further attracting enterprises to internal association. A similar structure is observed in Zaporizhzhia (Figure 2). 28. Kommunar Corporation 3. FED 39. AV metal group 8. ROSS 38. Ukrenergoatom 10. Ukrainian weighing company 34. Velton Telecom 25. Motor Sich 12. Thermo-Engineering Ltd 35. PrivatBank 43. Metinvest 48. Raiffeisen Bank
As can be seen from Fig. 2, the first 6 levels are occupied by those enterprises that have prospects for clustering. Enterprise No. 25 “Motor Sich” has the necessary elements of interaction in the cluster, which can enable productive cooperation between customers and suppliers. By involving active enterprises of the first levels of the hierarchy in cooperation, it is possible to realize the prospects for interaction between cluster members on the terms of benefits for everyone, which will not only expand the list of participants, but also allow constructively increasing the level of dialogue between them. It is necessary to assess the level of public interest in these enterprises in the regions. Google Trends was used as one of the computational linguistics tools to complete the task.
Google Trends [14] is an Internet application that allows you to determine the frequency of search queries (terms) in different parts of the world and in different languages. In addition, the application allows the user to compare, analyze data over a period of time, search for non-obvious topics related to the query. The application is a tool of computational linguistics, because it can be used to determine the filtering of search results using punctuation. Google Trends search results are displayed in interactive graphs that show the current situation of the top queries. This application helps users to delve into scientific issues for free, collect large amounts of data and track any changes in the dynamics of interest in the concepts. Unlike other software products, such as AWStats, Matomo, which have similar functions, Google Trends has a simple interface and remains the most popular for users in Ukraine.
Google Trends algorithms are based on determining the ratio of the number of term requests to the total number of queries on this concept. The obtained data are automatically translated into a 100point scale, which assigns the highest value of 100 and equates with it all other data, giving them points from 0 to 99 [15, 16, 17]. It should be noted that Google does not display rare, single, unpopular requests of one user in a short period of time. On the contrary, query statistics can provide results for phrases that include words in any order or in combination with other words [18, 19]. Thus, the capabilities of Google Trends and its functions make it possible to analyze the level of interest in the regions previously obtained as priority enterprises for cooperation (see Fig. 1, 2).
To determine the correspondence of the cooperative perspective to the modern public interests of business, a study was conducted to compare certain subjects of Kharkіv with each other. The analysis was carried out for three enterprises in the region in the last year (12 months). Search results in Google Trends are shown in Figure 3. As shown in Figure 3, the pre-defined cooperation links fully reflect the public Internet user requests. “Kommunar Corporation”, “FED” are Kharkiv enterprises representing the level of interest in the region in the proportion of 50/50, which demonstrates the demand for their activities by location. In addition, the neighboring Dnipropetrovsk region and the capital city of Kyiv are interested. “AV metal group”, identified as a promising cooperation partner for the cluster in the region, really corresponds to such a mission, because the demand for its products covers 10 regions of the country, which can be said about the existing partnerships between companies in these regions. So, the group of enterprises represented reflects the real state of development of future prospects for interaction between certain enterprises in Kharkiv.
A similar study was conducted according to three enterprises in Zaporizhzhia (Figure 4).
As can be seen from Figure 4, “Motor Sich” is a model of public interest in all regions of the country, which reflects the prospects of partnership for such a company in the sectoral cluster. “Metinvest” is most interested in Donetsk and Dnipropetrivsk regions. Instead, in the search space for any variations of “Thermo-Engineering Ltd” is absent, which indicates inconsistency with the requests of the companies themselves in the survey. Thus, it is doubtful to take into account “ThermoEngineering Ltd” when forming cooperation prospects, because the lack of mentions in Google Trends emphasizes the low level of interest in the company's products. Moreover, there are several companies in Ukraine with a commercial name that includes the words "thermo" and "engineering", and a significant number in the world. Cooperation in the event of such a discrepancy may create additional risks in the formation of associations.
The analysis of Google Trends made it possible to assess, using computational linguistics, the real state of possible interaction for future cooperation based on the demand for the company's products in the network, but this statistics does not guarantee the quality of the association, since it does not take into account the risks of this process. To determine the qualitative indicators of cooperation and their compliance with the results obtained, Google Trends suggests using a scoring matrix approach for assessing the corresponding risk, which includes elements of linguistic scaling.
To determine the criteria for assessing the risk of cooperative ties between enterprises, it is necessary to consider the stages of the formation of a cluster structure that precede the merger. In the scientific literature, said the risks of cooperation arise at the stages: choosing partners, planning and implementing a cluster project, assessing the cluster synergy, making a decision to form a cluster. Each stage is accompanied by risks inherent only in certain processes of interaction between the two partners. In the study of O. Palyvoda [20] The system of certain risks within the presented clusters (Table 1), requiring linguistic generalization in order to be further used in the processes of direct evaluation of compounds. The list of risks is not exhaustive and can be constantly updated depending on the current aspects of economic development, but any changes will not significantly affect the logic of generalization.
Specific risks – the risk of – the risk of unilateral use of - the risk of collecting - risk of cluster choosing partners cooperation; inaccurate information identification; with incompatible - the risk of dependence on about the results of - risk of business models; partners; cluster interaction; assessing the – the risk of – the risk of losing technological - the risk of using specialization misunderstanding innovations; inappropriate methods and prospects the partners place – risk of loss of commercial for assessing the of the cluster; and role in the information and organizational effectiveness of cluster - the risk of cluster network; innovations; interaction; assessing the – the risk of – risk of long-term decision- - lack of qualified existing mismatch of moral making; analysts on cluster economic values. – the inability to alternatively cooperation. conditions of use the resources involved in the cluster; the cluster project; - the risk of – the risk of additional costs for finding the organization of common communications. interests.
incompatibility
The unified matrix approach is a risk scoring method that takes into account not only the impact of cooperative risks, but also the likelihood of such risks occurring if agreements between partners are not fulfilled. In 2020, the Ministry of Finance of Ukraine approved the methodological manual "Riskoriented planning of internal audit activities" [21], which actually provides recommendations for assessing any risks of enterprises in probability and impact, taking into account linguistic means of visualization. To achieve the objectives of the study, it is proposed to adapt the process of assessing the cooperative risk to the conditions of this methodology. So, taking into account the criteria of cooperation risk defined above, the level of its influence on the prospect of clustering (low, medium, high and very high) is determined. This unified way, the level of risk is determined by assigning an appropriate score to it (Table 2).
As can be seen from Table 2, the discrepancy in the criteria makes it possible to assess the level of impact of cooperation risk on the future partnership, thus presenting it in matrix form by assigning it to a certain quadrant. For clarity of the matrix, it is proposed to supplement it with criteria for the probability of occurrence of the risk of cooperative relations of enterprises. Since the level of risk exposure is determined by 4 components, it is proposed to introduce 4 criteria (Table 3).
point 1 2
As can be seen from Table 3, each level of probability meets certain criteria, which are assigned an appropriate score, which allows you to adjust the previous estimate, which is presented in Table 2. Overall risk assessment is determined by multiplying the impact assessment and probability estimate as shown in the matrix Ministry of Finance of Ukraine [21] (Figure 5).
As can be seen from Figure 5, risks from levels 1 to 4 are marked in blue and are considered "low" (or acceptable), from levels 6 to 9 - "medium" risks (yellow), with levels from 12 to 16 - " high risks "or" very high "(red and orange). This matrix approach allows for a two-level assessment of the cooperation risk of enterprises, taking into account its impact and probability of occurrence for the organization. Further, to test the effectiveness of the method, it is proposed to conduct a risk assessment for pre-selected enterprises that are promising for cooperation and membership in existing regional clusters.
(experiment)
Assessment of the impact and probability of assigning points can be done in different ways, but each of them is based on an individual survey as a tool for collecting information. Thus, in order to assess the risk of the priority Kommunar Corporation as enterprise of the potential cluster of Kharkiv, an oral survey of representatives of the enterprise was conducted, the results of which are shown in Table 4.
incompatibility
on communication
assessment of the cooperation confirms the good faith intentions of the contractors
Kommunar Corporation partners have identical interests that can be expressed in joint projects, although the economic conditions for the development of cooperation are unstable
similar business does not provoke models with the dependence, but members of the cluster only creates (Zaliznychavtomatika, additional UKRDIPROVAZHMASH, opportunities for etc.), they have the earnings for same role in terms of Kommunar importance at the level Corporation, resource of state strategies outflow, including financial will not happen, because the cooperation will be carried out on the terms of mutual guarantees in the cluster
Low (
As can be seen from Table 5, the probability of occurrence of different evaluation criteria has a discrepancy, covering the range from 1 point to 3. Opinions of experts in the survey coincide. According to them, the most probable aspects that negatively affect cooperation may be the fleeting economic conditions in the country and the dishonesty of partners in providing reliable information on the results of cooperation.
Thus, by multiplying the two tables 4 and 5 in pairs, you can determine the total number of points by criteria and the quadrant of the matrix of risk assessment of cooperation (Table 6).
As can be seen from Table 6, the overall assessment of the level of risk indicates the possible cooperation of Kommunar with members of the Kharkiv regional cluster. Moreover, the low level of 3 of the 4 criteria suggests that cooperation is mutually beneficial for both Kommunar and their promising partners. The only obstacle remains the unstable economic situation, which could significantly change the course of the partnership. However, Kommunar's readiness to cluster under such conditions is obvious, which allows further promotion of this idea to the media space of potential participants in the cooperation.
The results of the assessment of the level of risk of cooperation of «Kommunar Corporation» with promising partners of the Kharkiv regional cluster completely coincide with the conclusions on public interest requests in Google Trends, which is a tool of computational linguistics that helps integrate statistics to potential users. That is, we can conclude that Google Trends is a relevant tool for assessing the interest in cooperation, which can reflect the main trends in the relevant industry selected for analysis. The demand for certain requests characterizes the real-time state of affairs for companies that aim to expand their own prospects for partnership.
The results obtained to determine the risks of cooperative relations of enterprises were based on the technology developed by the authors of processing quality information with widespread use of methods of using linguistic information, which is important for forming the preconditions for decision-making in the economy.
The practical significance of this research lies in the developed approach and the actual experiment based on multi-step transformations: using the online Google Forms for the survey, applying the hierarchy method, determining interest in Google Trends tools, building linguistic scales, combining the results of each step to confirm economic decision in the potential construction of cooperative ties of the cluster at the regional level in the priority development zone of Industry 4.0 of the industrial region of Ukraine.
The developed technology in further research can be used to achieve the goals of computational
linguistics in terms of improving the processing of user requests on the Internet and to conduct a more
in-depth analysis of the terms that precede any research in economics. To more deeply determine the
state of enterprise cooperation, research can be scaled up by regions of Ukraine or in the global
European space. Also, to understand the place of each studied object in the risk zone and to determine
possible trends in the transition from one risk zone to another, it is possible to supplement the study
with similar calculations in the dynamics.
[4] D. Shmatkov, S. Hlibko, K. Tokarieva, J. Cachón Zagalaz, On the question of why copyright
cannot be synonymous with intellectual property within digital competence frameworks, On the
question of why copyright cannot be synonymous with intellectual property within digital
competence frameworks. Revista La Propiedad Inmaterial, 32 (2021) 215–231. doi:
https://doi.org/10.18601/16571959.n32.07.
[5] V. Vysotska, Information technology for internet resources promotion in search system based on
content analysis of web-page keywords. Radio Electronics, Computer Science, Control (