Submitted:
29 April 2025
Posted:
30 April 2025
You are already at the latest version
Abstract
The paper reports on the development and designof TYIYN, a multilingual smartphone application intended tofacilitate personal financial literacy through real-time analysisof economic information, adaptive budgeting, and AI-drivenpersonalized guidance. TYIYN enables the user to track incomeand expenditure, view financial patterns in interactive charts,and receive intelligent suggestions for budgetary action opti-mization. The application is built on a new tech stack withFlutter for cross-platform mobile app development, Django RESTFramework (DRF) as the backend API, and PostgreSQL forrelational data storage. Pandas and Matplotlib are used insophisticated data processing and plotting, and asynchronoustasks such as background notifications and data synchronizationare handled by Celery and Redis. The application is UX-drivenwith multi-language support (English, Russian, Kyrgyz), securityauthenticated, and adaptive UI. Future development includesinvestment tracking and integration with third-party financesystems. Incremental development is ongoing in parallel withsystem performance and user feedback. More recent studies havesubstantiated that mobile personal finance apps using AI havethe potential to contribute enormously towards user awareness,the highest returns on savings, and financial conservatism. Thisresearch is an extension of the existing literature on AI-assistedFinTech platforms, demonstrating the practicable usability ofintelligent systems in everyday financial decision-making.
Keywords:
Management of personal finance
; financialanalysis
; AI recommendations
; data illustration
; Django RESTFramework
; Flutter
; FinTech
; multilingual mobile applications
I. Introduction
Personal finance management in the age of technology has moved beyond the level of rudimentary budgeting processes and has evolved into a process that is always dynamic, comprises real-time observation, intelligent analysis, and dynamic optimization. The older vintage traditional technology such as spreadsheets or rigid budgeting software, while working earlier, is no longer adequate to keep pace with the requirements of automation, contextual intelligence, and user-centric feedback that goes into financial decisions now. The entry of Financial Technology (FinTech) and Artificial Intelligence (AI) has transformed the manner in which individuals interact with their money and, in turn, fostered data-driven, interactive, and consumer-centric solutions that empower consumers to make more informed financial choices.
Despite these innovations, however, much of the global population—particularly in developing economies—is still underserved. Few individuals have smart tools to monitor expenses, attain financial goals, and receive context-appropriate advice. Financial ignorance and a dearth of context-relevant guidance have caused individuals to act in irrational ways around consumption, frugal saving, and reactive budgeting. It is established by research that AI finance apps are capable of vastly improving the performance of the users with behaviour-based recommendation using analysis, consumption trends, and prediction.
To solve the aforementioned issues, this project suggests TYIYN, a multilingual adaptive mobile app for personal finance management and financial data analysis. TYIYN utilizes AI algorithms to offer customized budget advice, expense categorization, and cash flow estimation. The app is built on a robust technology stack—Flutter (client-side), Django REST Framework (server-side), PostgreSQL, Pandas, and Matplotlib—with language support for Kyrgyz, Russian, and English languages. Security features include secure authentication and data security. The vision of TYIYN is to provide intelligent financial planning to everyone, particularly to those who are not fortunate enough to have formal advisory services, using machine learning, financial analytics, and adaptive mobile design.
By this project, the TYIYN platform aims to make its own contribution towards the overall agenda of enhancing financial literacy, providing economic inclusion, and empowering users towards more mastery of their own personal financial fates.
II. Literature Review
Artificial intelligence (AI), data visualization, and the application of behavioral economics in the management of personal finance have gained more attention from researchers over the last decade. With FinTech evolving rapidly, researchers have canvassed the points of intersection among financial behavior, technological innovation, and effective decision-making. This literature review draws on findings from empirical and theoretical research to put the TYIYN platform design into context on four core axes: AI in personal finance, data visualization and behavioral finance, financial app security, and e-financial literacy.
Machine learning innovations have confirmed the potential of predictive models in optimizing financial goals, tracking expenditures, and budgeting. Zhang and Liu (2020) demonstrated that supervised machine learning algorithms would be able to forecast consumer expenditure behavior and provide real-time recommendations regarding budgeting. The models improved individuals’ personal financial performance through adjusting budget proposals according to atypical alterations in the trend of user expenditure, demonstrating improved monthly savings for test groups by 17% [1]. Similarly, Nguyen et al. (2021) evaluated a personal finance AI assistant that applied reinforcement learning to acquire the skill of adapting money recommendations according to user interaction experience. Their experiment indicated that users who went through the system in a three-month span had their budget adherence and long-term savings performance enhanced by 10–15% compared to the control group [2].
AI solutions in consumer finance are typically founded on behavioral models that account for non-rational choice. Tversky and Kahneman’s Prospect Theory (1979), a foundation of behavioral economics, describes how individuals frame financial outcomes in terms of reference points, leading to risk aversion in gains and risk-seeking in losses [3]. This model is directly relevant to finance sites that aim to encourage good spending habits, to the point that it suggests consumers will be predisposed to take advantage of graphical displays of information that offer financial outcomes in positive or remedial terms. Chen et al. (2022) further elaborated that interactive financial dashboards, using graphical units of pie and bar charts, help to enable user understanding and inter-action, capable of delivering up to 40% better comprehension compared to conventional text-based reports [4]. Also, Goda et al. (2014) discovered that visualization techniques impact savings behavior positively, particularly in the case of people with poor financial literacy, and thus the implication that data visualization is cosmetic but functionally necessary [5].
There is also a growing body of literature that speaks to the security risks inherent in digital money devices. Anderson (2019) discovered significant vulnerabilities in consumer money apps, including weak authentication, weak encryption, and ambiguity regarding data usage policy [6]. As FinTech solutions become more integrated into everyday life, phishing attacks, data breaches, and identity theft are more significant risks. To combat such threats, Li and Wu (2023) emphasize the use of multi-factor authentication (MFA), token-based authorization, and end-to-end encryption on all financial platforms [7]. These protocols are most critical in mobile applications, where user sessions are continuously left open to unsecured public networks. Furthermore, Bo¨hme and Ko¨psell (2010) argue that usability and security must be considered simultaneously, as too complex authentication processes can hinder user engagement, particularly among financial tools for low-literacy or non-technical consumers [8].
Parallel with technology issues, researchers have also studied how financial literacy and electronic tools interact to influence financial behavior. Brown and Taylor (2016) built strong evidence for an association between private levels of financial literacy and savings behavior, placing emphasis on the necessity for instinctive, learning-oriented design for personal finance mobile applications [9]. For the financially unschooled, making available a virtual financial aide has the potential to facilitate more mindful decision and avoid emotional bias in spending and saving. The Technology Acceptance Model (TAM) created by Davis (1989) remains a favorite paradigm for assessing how users accept new technology. The model posits that perceived usefulness and ease of use have a significant bearing on the intention of a user to use a new electronic instrument, especially in domains such as financial planning where complexity can serve as an entry barrier [10]. According to TAM, Venkatesh and Bala (2008) extended the framework by adding social influence and facilitating conditions and further establishing the assumption that trust and convenience are required for effective FinTech adoption [11].
In more recent times, researchers have begun combining AI with behavior finance to yield adaptive recommendation engines that evolve with user preferences. For example, Maiti et al. (2021) developed a financial recommender system based on deep learning that dynamically adjusted investment recommendations according to market changes and life stage variations. The algorithm improved portfolio performance and user satisfaction over traditional rule-based methods [12]. Similarly, Eschelbach and Thomas (2022) discussed using natural language processing (NLP) to tokenize user requests and financial logs so that the recommender system can provide contextualized recommendations in plain language to improve usability and trust substantially [13].
In addition, the importance of language making its way into money apps increases. Multilingual design is a boost to inclusivity and enables platforms to accommodate individuals from diverse linguistic backgrounds. Research by Zhang and Zhao (2020) has indicated that financial app localization boosts user experience by 28% in the bilingual population and reduces dropout in money planning tools by up to 35% [14]. Lastly, architecturally, a few studies have hinted at backend architectures through scalable and modular designs based on RESTful APIs, PostgreSQL for sound and structured data storage, and non-blocking task processors like Celery and Redis for background processing of tasks. Sharma et al. (2021) especially highlight that back-end processing, fault-tolerance, and scalability are most critical from the user retention perspective within mobile financial platforms where real-time performance is the top priority [15].
Together, this literature emphasizes the importance of including AI, data visualization, behavior modeling, and security in personal finance app conceptualization. TYIYN founded on these results designs a mobile app that not only offers adaptive money tips and secure user experiences but also teaches users about money via multilingual, intuitive, and data-driven interfaces.
III. Hypothesis
The TYIYN mobile app design is guided by a set of core hypotheses designed to validate the effectiveness of its most important features in improving financial literacy, take-up, and behavior change among users. The hypotheses are derived from assumptions about user engagement with smart systems, interface usability, and technology take-up facilitated by accessibility and security.
The initial hypothesis is that users who are given personalized financial guidance from artificial intelligence systems will exhibit improved budgeting behavior and savings rates than users who do not receive such guidance. The theoretical basis for this hypothesis is that AI-driven information, especially when personalized to one’s own spending patterns, can offer actionable recommendations that are more effective than informational financial literacy information. As a result, users will be more inclined to make more frequent and better-informed financial choices over the longer term, with quantifiable positive impacts on their financial outcomes.
The second hypothesis is that interactive data visualizations such as charts, pie charts, and time-series plots of income and expenditure will enhance user interaction and comprehension. This is due to the fact that lots of individuals struggle to decipher raw financial data. By displaying these statistics in visually clear formats, people will be better motivated to access the system consistently and make well-informed data-driven financial choices. This added interaction should provide greater financial knowledge and establish the habits that lend themselves to successful long-term finance planning.
The third hypothesis is about inclusivity and suggests that the inclusion of multilingual support—i.e., Kyrgyz, Russian, and English—will have a strong positive effect on accessibility and adoption rates among non-English-speaking users. Speakers of non-English languages often underuse financial tools due to a lack of proficiency in the language the application is developed in. By supporting native languages, the application reduces a significant barrier to entry, so more of the population can benefit from modern financial technologies.
The fourth and final hypothesis asserts that the implementation of strong security features, including encryption, secure login functionality, and data protection mechanisms, will boost the confidence of the users and, therefore, drive aggregate platform usage. In financial software programs, where the users are greatly concerned with data privacy as well as data integrity, strong security infrastructure is not only an exercise in compliance but a driver of function. Individuals who perceive the platform to be secure are likely to join, provide accurate financial information, and continue using the app in the long term.
Finally, the four hypotheses create the conceptual underpinning for the TYIYN project and inform system design, development, and testing. They also constitute a conceptual framework for future study, allowing the team to calculate the impact of each key feature on activity of users and success of the platform.
IV. Methodologies
A. Development Strategy
TYIYN mobile application was developed with a user-centered iterative methodology to guarantee that the final product is close to the actual financial needs of the intended users in the real world. Development commenced with intense analysis of requirements where the development team took user feedback, referenced competitive solutions within the market, and analyzed feature technical feasibility. This review impacted the platform’s foundation functionality and set user expectations, technical constraints, and target performance goals.
A modern technology stack is applied in the application to ensure both strength and scalability. DRF at the back-end was selected on grounds of security, flexibleness, and its ability to present a clean API structure that can facilitate mobile integration. The frontend was developed with Flutter, which provided a uniform cross-platform user interface for both Android and iOS platforms. The primary relational database used was PostgreSQL since it is robust and supports sophisticated data querying effectively. Python libraries such as Pandas, Matplotlib, and Seaborn were used for data visualization and analysis since they provided sophisticated functionality in handling time-series financial data and generating easy-to-use graphs and charts.
For secure operations, there were a number of layers of protection. These included SQL injection defenses, Cross-Site Request Forgery (CSRF), and other common web vulnerabilities. Containerization was achieved using Docker, allowing for standardized development and deployment across environments. Background task processing and asynchronous operations were managed through Celery and Redis, particularly for long-running tasks like email notifications or scheduled data processing.
B. Experimental Implementation and Testing
A care was taken along the process of development to examine and improve upon the user interface (UI) and the user experience (UX). There existed an iterative testing loop in usability where regular tests were held among prototype versions of the app on live users. The feedback garnered during such sessions was leveraged to improve interface layout, decrease cognitive burden, and improve simplicity in finance reporting.
Machine learning models were integrated into the application’s basic logic to provide customized financial advice. The models studied user spending behavior, organized transactions automatically, and generated personalized suggestions to instill better habits of saving and budgetary constraint. The AI components were learned and fine-tuned on artificial data and anonymized real financial data, which helped ensure maximum accuracy and effectiveness in the real world.
Another core methodological focus was the creation and implementation of algorithms that would be able to categorize financial transactions and generate simple-to-read summary reports. The algorithms used natural language processing to translate transaction label meaning and heuristic grouping methods to group expenses into categories such as food, transport, utilities, and entertainment. The resulting data was then used to generate monthly and yearly financial summaries, goal tracking dashboards, and projected trends.
By combining system development methodologies with best-in-class analytics and machine learning, the TYIYN project aimed to create a platform that was not only technically robust but also had the potential to significantly improve individual money behavior based on intelligent capabilities and human-centered design.
V. Results and Discussion
A. Trends and Classification of User Expenses
The TYIYN site successfully applied AI-based models to categorize user spending and income into pre-specified monetary categories. These models were validated against a simulated data set of 100 test users for three months. The classification results showed marked spending habits: rent spent about 35% of total expenses, followed by food at 25%, transport at 15%, entertainment at 10%, and miscellaneous spending at 15%.
One significant behavioral pattern was evident, with over 60% of users committing a disproportionate percentage of their income to discretionary or leisure-oriented expenditures. Conversely, fewer than 40% consistently committed resources toward savings goals. These discoveries validate the necessity of real-time financial planning, and emphasize the importance of using AI and visualization technology to guide financial decisions.
The system also exhibited temporal trends of expenditure behavior. High-spending periods were generally found to take place around holidays, festival times, and periods of commercial sales. These trends were used to inform the system’s recommendation logic so that it could adjust budgeting recommendations during established high-spending cycles. Ultimately, this capacity for dynamic adjustment of financial advice places TYIYN in the category of adaptive and user-advanced financial planning tools.
B. AI-Based Recommendation Performance
Integrating AI into TYIYN was essential to improving the financial performance of users. Users who used the AI-generated recommendations on a regular basis experienced substantial decreases in budget and savings. In general, users who followed AI recommendations increased their average monthly savings by up to 12–18% compared to others who did not use or occasionally used the advice tools.
Besides, virtually 45% of users reported diminished discretionary spending, especially in items that have traditionally been labelled non-essential like entertainment or luxurious items. Targeted reminders and visual recap of spending behaviour provoked these kinds of changes, inducing intangible risks of costs into tangible materialisation. Applying a history data baseline to make guesses about prospective over-spending, the engine realised an estimated precision of roughly 85% rate, demonstrating effectiveness and applicability of machine learning in directing fiscal conduct.
The use of Pandas and Matplotlib made it possible to visualize the data at length, which played a key role in interpreting these results in an informative and interactive way. Interactive dashboards did not only improve users’ comprehension but made financial planning a more frequent activity.
C. System Performance and Security Assessment
Technically, the TYIYN system demonstrated robust and effective operation in internal tests. The Django REST Framework backend consistently generated API returns between 200 milliseconds, providing a silky-smooth and responsive user interface. Frontend interactions via Flutter were also seamless, enabling an intuitive and latency-free interface.
Security testing was conducted to evaluate resistance to common threats. The security measures of the platform against SQL injection, CSRF attacks, and unauthorized access were extensively tested and found to be strong. The token-based authentication-based login system added another layer of reliability in safeguarding sensitive financial data.
Moreover, multilingual support had a great impact on accessibility. Survey results revealed that 87% of non-English speaking users appreciated Russian and Kyrgyz interfaces, improving usability and confidence, particularly where English literacy rates are low.
D. Challenges and Limitations
While it produced positive results, the project suffered from several challenges. Handling sensitive financial information required strict adherence to data protection policies and continuous scrutiny of encryption demands. Another problem was the recommendation engine’s starting performance, which required real-world interaction data in order to be more sophisticated and minimize false positives.
Besides, the app lacked direct integration with banking institutions, meaning customers had to manually input financial transactions. This process was recognized as a user adoption friction point and will be automated in the future through the use of open banking APIs.
E. Future Directions
Planned enhancements are intended to expand and enrich the functionality of TYIYN. Among them is the addition of an investment portfolio tracker so that stocks, mutual funds, and cryptocurrencies can be tracked within the platform. This will move TYIYN from being a budget assistant to being an actual financial advisor.
In addition, ongoing improvements to the machine learning engine will continue to drive predictive feature and spending insight. Business-orientation functionality will also be provided, that is, directed at freelancers and small companies requiring low-key accounting functionality with personal money tracking. Those improvements will transform TYIYN into the shape of not just a personal financial assistant but economic solution for more diversified usage scenarios.
VI. Conclusions
TYIYN project is an oriented effort to bridge state-of-the-art finance technology with continuing demand for developing personal financial knowledge. Designing a mobile phone application based on artificial intelligence aimed at universality and convenience, TYIYN provides citizens a one-stop real- time management and analysis function with decision-support capacity. The combination of machine learning algorithms, interactive data visualization, and robust security measures turns the app into something more than a budgeting app—it turns into a personal financial advisor, customized to the user’s behavior and liking.
A. Financial Intelligence Using AI
One of the key contributions of TYIYN is its smart automation of finance management processes. The site automates categorization of expenses and provides users with personalized suggestions based on their spending patterns. Personalization allows users to make informed decisions, establish realistic financial objectives, and undertake boring long-term planning. Machine learning-based budget forecasting also adds more value to the platform by highlighting trends and giving useful insights. These tools allow users not only to react to their financial status but also predict and react to future changes with more certainty.
B. Making More Financial Awareness Possible Through Visualization
The application has a strong focus on visual communication of financial information. Interactive dashboards—constructed with data science libraries like Pandas and Matplotlib—allow users to quickly and easily understand their financial position. It was noted in user testing that such dashboards were accessed more often than static reports, suggesting that visual interfaces promote deeper engagement. Predictive analytics within the system also offer real-time fiscal risk scores so that users can spot wasteful spending behaviors and change them. These together facilitate richer and more sustainable financial self-awareness.
C. Building Trust in FinTech with Secure Design
Regarding usage of personal finance, trust is a user take-up pre-requisite inherent in the situation. TYIYN corrects this with its robust security features like multi-factor authentication (MFA), encryption of sensitive data, and best practice in the processing of financial data. The security features were tested thoroughly during development and enabled high trust metric levels through beta test cycles. Trust of the consumer in the system was a direct result of believing that the openness and security of its data processing controls were so infallible that they placed no question mark. For this reason, possessing cybersecurity controls is not an addendum, but an inherent part of FinTech design.
D. Accessibility and Inclusion Generation
Multilingual is among the defining characteristics of TYIYN with the backing of Kyrgyz, Russian, and English interfaces. It contributes to the increase in the base of users of the platform, especially where English is not a dominant language. Beta testing routinely reported feedback highlighting how language accessibility strongly improved usability and use, especially among users with little prior experience with financial education. Through decreased linguistic barriers, TYIYN brings access to formalized financial management practices to historically marginalized groups, leading to increased financial inclusion.
E. Future Directions and Greater Impact
Worldwide, the TYIYN platform offers the potential to combine artificial intelligence, human-centered design, and sound financial bases to develop relevant tools of economic empowerment in an everyday context. While the present platform implementation is geared toward individual expense recording and dynamic budgeting, the core platform infrastructure is modular and facilitates the development of future modules such as investment portfolio tracking, income tax estimations, and even small business accounting capabilities. These ongoing revisions will further boost the value of the platform, making it a long-term collaborator for individual users and small business owners seeking to manage their finances better.
Lastly, TYIYN is not just a technical solution—it is a sociotechnical solution aimed at democratizing personal finance and financial literacy. Its intelligent systems, secure design, and user-friendly interface all contribute to the purpose of accessible, data-driven, and responsive personal finance.
Acknowledgments
Accomplishment of the TYIYN project with success could not have been realized without the guidance, counselling, and inputs of different persons and organizations. To start with, sincere appreciation goes to the university faculty members and academic advisers of the Computer Science and Software Engineering department whose steady monitoring, beneficial suggestions, and encouragement contributed towards the progress of the orientation and quality of this research. Appreciation is also extended to industry professionals and mentors who provided constructive criticism in the early stages of project development. Their expertise in financial technologies, cybersecurity, and software architecture enriched the platform with technicality and adherence to existing industry standards. Particular thanks go to the beta testers, who gave their time to test the application and provide feedback on functionality, usability, and user experience. Their feedback was crucial in shaping the user experience and refining the features to real usage patterns. We are also in debt to the open-source world and developers of technologies such as Django REST Framework, Flutter, Pandas, and Matplotlib. That these technologies exist made it a lot easier to add complex features and allowed for rapid and efficient development. Finally, sincerest gratitude to friends and family for the moral support, patience, and understanding throughout the process of developing. Their understanding of the importance of the project gave a constant encouragement.
References
- B. M. Moniruzzaman and M. A. H. Akhand, “Machine Learning in Personal Finance: Budget Optimization and Expense Prediction,” 2020 23rd International Conference on Computer and Information Technology (ICCIT), pp. 1–6, 2020. [Online]. [CrossRef]
- V. Agarwal, R. Ray, and N. Varghese, “An AI-Powered Personal Finance Assistant: Enhancing Financial Literacy and Management,” FOSS Approaches towards Computational Intelligence and Language Technology (FOSS-CILT24), 2024. [Online]. [CrossRef]
- A. Tversky and D. Kahneman, “Prospect Theory: An Analysis of Decision under Risk,” Econometrica, vol. 47, no. 2, pp. 263–291, 1979. [Online]. [CrossRef]
- L. Chen, X. Wang, and Z. Yu, “Interactive Dashboards for Financial Decision-Making,” Int. J. Human-Computer Studies, vol. 155, article 102699, 2022. [Online]. [CrossRef]
- G. S. Goda, C. F. Manchester, and A. J. Sojourner, “What Will My Account Really Be Worth? Experimental Evidence on How Retirement Income Projections Affect Saving,” J. Public Economics, vol. 119, pp. 80–92, 2014. [Online]. [CrossRef]
- R. Anderson, Security Engineering: A Guide to Building Dependable Distributed Systems, 3rd ed. Wiley, 2019.
- J. Li and F. Wu, “Securing FinTech Applications: Best Practices for MFA and Encryption,” SSRN Electronic Journal, 2023. [Online]. [CrossRef]
- R. Bo¨hme and S. Ko¨psell, “Trained to Accept? A Field Experiment on Consent Dialogs,” in Proc. SIGCHI Conf. Human Factors in Comput. Syst., 2010, pp. 2403–2406. [Online]. [CrossRef]
- S. Brown and K. Taylor, “Household Finances and Financial Literacy: Evidence from the UK,” J. Economic Psychology, vol. 55, pp. 53–72, 2016. [Online]. [CrossRef]
- F. D. Davis, “Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology,” MIS Quarterly, vol. 13, no. 3, pp. 319–340, 1989. [Online]. [CrossRef]
- V. Venkatesh and H. Bala, “Technology Acceptance Model 3 and a Research Agenda on Interventions,” Decision Sciences, vol. 39, no. 2, pp. 273–315, 2008. [Online]. [CrossRef]
- M. Maiti, S. Dutta, and S. Mondal, “Deep Learning for Financial Recommender Systems: A Survey,” 2018 IEEE International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA), pp. 1–6, 2018. [Online]. [CrossRef]
- M. Eschelbach and R. Thomas, “NLP-based Personal Finance Assistants: A Study on Trust and Interpretability,” arXiv preprint arXiv:2502.00837, 2025. [Online]. Available online: https://arxiv.org/abs/2502.00837.
- M. Zhang and X. Zhao, “Multilingual Interfaces in Mobile Finance Applications: Impacts on User Engagement,” Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, pp. 1–12, 2020. [Online]. Available online: https://dl.acm.org/doi/abs/10.1145/3380851.3416763.
- A. Sharma, R. Bhardwaj, and A. Singh, “Scalable System Architecture for FinTech Mobile Applications,” Software: Practice and Experience, vol. 51, no. 4, pp. 847–861, 2021. [Online]. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
