The integration of artificial intelligence in veterinary medicine represents a paradigm shift toward precision livestock management. This study presents the development and implementation of an intelligent expert system designed to evaluate the therapeutic effects of Naftalan oil on ovine gastric function. Through systematic analysis of dose-response relationships across four treatment groups (5, 10, 15, and 20 ml) compared to control conditions, we demonstrate significant improvements in gastric pH levels, rumination duration, and feed intake capacity. The expert system architecture encompasses a comprehensive knowledge base, inference engine, user interface, and adaptive learning module, achieving 95% accuracy in dosage recommendations while reducing decision-making time by 70%. Our findings reveal that the 20 ml dosage produces optimal therapeutic outcomes, with gastric pH increasing by 1.5 units, rumination time extending to 7 hours, and feed intake capacity improving by 20%. Hematological parameters showed concurrent improvements, including 7% increase in erythrocyte count, 8% elevation in hemoglobin levels, and 10% rise in leukocyte count. This research establishes a foundation for intelligent decision support systems in veterinary practice, demonstrating how traditional therapeutic knowledge can be systematically integrated with modern computational approaches to enhance animal health outcomes [1-3].

Keywords: Naftalan Oil; Expert Systems; Ovine Medicine; Gastrointestinal Function; Veterinary Informatics; Decision Support Systems

The convergence of traditional veterinary medicine with advanced computational intelligence marks a transformative era in animal healthcare. As practitioners increasingly recognize the limitations of empirical approaches to complex physiological interventions, the development of evidence-based decision support systems has emerged as both a necessity and an opportunity. This evolution is particularly significant in the context of natural therapeutic agents, where the intricate relationships between dosage, physiological response, and individual animal characteristics demand sophisticated analytical frame works [4-6]. Naftalan oil, with its unique hydrocarbon composition enriched with nitrogen and sulfur compounds, represents a compelling case study in this intersection of traditional therapy and modern Analysis [7-9]. The therapeutic potential of this naturally occurring substance has been recognized for decades, yet its systematic application in veterinary medicine has been constrained by the absence of standardized protocols and evidence-based dosing guidelines. The complexity of gastrointestinal physiology in ruminants, coupled with the multifaceted pharmacological profile of Naftalan oil, creates a challenging landscape for clinical decision-making that traditional approaches struggle to navigate effectively [7-9]. The emergence of expert systems in veterinary medicine addresses these challenges by providing a structured framework for knowledge representation, inference, and decision support. Unlike conventional diagnostic tools that rely primarily on pattern recognition, modern expert systems integrate multiple data streams, historical outcomes, and physiological models to generate contextually appropriate recommendations. This approach is particularly valuable in livestock medicine, where individual animal assessment must be balanced against herd-level considerations and economic constraints [5,10,11]. Our research addresses a fundamental gap in the current understanding of Naftalan oil's therapeutic applications through the development of an intelligent expert system specifically designed for ovine gastrointestinal health management [12-14]. The system represents more than a simple decision tree or lookup table; it embodies a comprehensive approach to knowledge integration that combines biochemical understanding, physiological modeling, and empirical validation. By systematically evaluating dose-response relationships and integrating these findings into a dynamic knowledge base, we provide practitioners with a tool that evolves with experience and adapts to new evidence. The significance of this work extends beyond the immediate application to Naftalan oil therapy. The methodological framework we present offers a template for developing expert systems across various domains of veterinary medicine, particularly those involving natural therapeutic agents where traditional pharmacological models may prove inadequate. The integration of adaptive learning capabilities ensures that the system remains current with emerging research and practical experience, addressing one of the primary limitations of static decision support tools.

The development of our expert system required a multifaceted approach that integrated experimental research, knowledge engineering, and system design principles. The foundation of this work rests on a carefully designed experimental protocol that evaluated Naftalan oil's effects across multiple dosage levels while simultaneously capturing the data necessary for robust system training and validation. Our experimental design encompassed five distinct treatment groups, creating a comprehensive dataset that spans the therapeutic range of Naftalan oil application. The control group provided essential baseline measurements, while the four treatment groups (5, 10, 15, and 20 ml dosages) allowed for detailed analysis of dose-response relationships. This systematic approach ensured that our expert system would be grounded in empirical evidence rather than theoretical assumptions, a critical factor in establishing clinical credibility and practical utility [13,15]. The selection of physiological parameters for monitoring reflected both the immediate effects of Naftalan oil treatment and the broader indicators of gastrointestinal health. Gastric pH levels served as a primary indicator of digestive function, while rumination time provided insights into the broader behavioral and physiological responses to treatment. Feed intake capacity measurements offered a practical metric directly relevant to livestock productivity, ensuring that our system addresses both health and economic considerations. Hematological analysis expanded our assessment beyond gastrointestinal function to encompass systemic effects of Naftalan oil treatment. The measurement of erythrocyte count, hemoglobin levels, leukocyte count, and total protein provided a comprehensive picture of the animal's physiological response, enabling the expert system to evaluate treatment effects holistically rather than focusing solely on local gastrointestinal changes [16-18]. The expert system architecture reflects current best practices in knowledge-based system design while incorporating innovations specific to veterinary applications. The knowledge base serves as the repository for all empirical findings, theoretical understanding, and practical guidelines related to Naftalan oil therapy [4,5,11]. This component required careful structuring to accommodate both quantitative dose-response data and qualitative clinical observations, ensuring that the system could integrate diverse forms of veterinary knowledge effectively. The inference engine represents the analytical heart of our system, implementing sophisticated algorithms that can process complex, multi-dimensional data to generate appropriate treatment recommendations [6,19]. Unlike simple rule-based systems, our inference engine incorporates probabilistic reasoning and uncertainty management, acknowledging the inherent variability in biological systems and the limitations of available data. This approach ensures that recommendations are appropriately qualified and that practitioners receive guidance on the confidence level associated with each suggestion. The user interface design prioritized accessibility and clinical workflow integration, recognizing that even the most sophisticated analytical capabilities are worthless if practitioners cannot easily access and interpret system outputs. The interface accommodates the reality of veterinary practice, where decisions must often be made quickly and in field conditions that may not support complex data entry procedures [4,20]. Perhaps most importantly, the adaptive learning module ensures that our expert system remains current and continues to improve over time. This component processes feedback from clinical applications, integrates new research findings, and refines recommendation algorithms based on accumulating experience. The learning capability addresses one of the fundamental challenges in medical expert systems: maintaining relevance and accuracy as knowledge evolves and practice patterns change.

The systematic evaluation of Naftalan oil's effects across multiple dosage levels revealed clear dose-response relationships that validate the therapeutic potential of this natural agent while providing the empirical foundation for expert system development. The progression of physiological improvements from the lowest to highest dosage levels demonstrated both the therapeutic window of Naftalan oil and the importance of precise dosing for optimal outcomes [8,13,14]. Gastric pH measurements provided compelling evidence of Naftalan oil's beneficial effects on digestive function. The control group maintained stable pH levels around baseline values, while treated animals showed progressive improvements corresponding to increasing dosage levels [15,21]. The 5 ml group demonstrated a modest 0.5 unit increase in gastric pH, suggesting initial therapeutic effects at this dosage level. The 10 ml group maintained similar pH improvements, indicating a potential threshold effect in the lower dosage range. However, the 15 ml group showed a notable increase to 1.0 unit above baseline, suggesting enhanced therapeutic activity at this dosage level. The 20 ml group achieved the maximum observed pH increase of 1.5 units, establishing this dosage as the most effective for gastric pH optimization within our experimental parameters. Rumination time analysis revealed parallel improvements that correlated strongly with gastric pH changes. Control animals maintained their baseline rumination duration of 5 hours, while treated groups showed progressive extensions of rumination time. The 5 ml group achieved a modest extension to 5.5 hours, with the 10 ml group reaching 6 hours. The 15 ml group demonstrated further improvement to 6.5 hours, while the 20 ml group achieved the maximum observed rumination duration of 7 hours. These findings suggest that Naftalan oil's effects extend beyond simple pH modification to encompass broader aspects of digestive behavior and function [8,22]. Feed intake capacity measurements provided perhaps the most practically relevant indicators of treatment success, directly linking physiological improvements to productivity outcomes. The control group showed no change in feed intake patterns, establishing a stable baseline for comparison [11,23]. The 5 ml group demonstrated a 5% improvement in feed intake capacity, indicating meaningful but modest therapeutic effects. Progressive dosage increases yielded corresponding improvements, with the 10 ml group achieving 10% enhancement, the 15 ml group reaching 15% improvement, and the 20 ml group attaining the maximum observed increase of 20% in feed intake capacity. Hematological analysis revealed systemic effects that complemented the gastrointestinal improvements observed in treated animals. Erythrocyte count increases ranged from 3% in the lowest dosage group to 7% in the highest dosage group, suggesting improved oxygen-carrying capacity and overall metabolic function. Hemoglobin level improvements followed a similar pattern, with increases ranging from 2% to 8% across the dosage spectrum. Leukocyte count elevations, ranging from 3% to 10%, indicated enhanced immune function without reaching levels that might suggest inflammatory responses. Total protein increases of 5% to 12% across dosage groups reflected improved metabolic status and potentially enhanced protein synthesis or absorption [17,18]. The expert system's performance in analyzing these complex, multi-dimensional datasets demonstrated the value of computational approaches to veterinary decision-making [19,20]. The system achieved 95% accuracy in generating appropriate dosage recommendations when tested against known optimal treatments, while reducing the time required for decision-making by 70% compared to traditional analytical approaches. These performance metrics validate both the quality of the underlying data and the effectiveness of the analytical algorithms implemented in the system. The integration of physiological and hematological data through the expert system revealed correlations and patterns that might not be apparent through conventional analysis. The system identified optimal dosage ranges not just for individual parameters but for comprehensive therapeutic outcomes, accounting for the complex interactions between different physiological systems. This holistic approach to data integration represents a significant advancement over single-parameter optimization approaches commonly used in veterinary medicine.

The successful development and validation of our expert system for Naftalan oil therapy represents more than a technological achievement; it embodies a fundamental shift toward evidence-based, data-driven veterinary practice. The clear dose-response relationships we observed across multiple physiological parameters provide compelling evidence for Naftalan oil's therapeutic value while demonstrating the importance of systematic dosing protocols in maximizing treatment efficacy [1,3]. The convergence of gastric pH improvements, extended rumination times, and enhanced feed intake capacity at the 20 ml dosage level suggests that this represents a true therapeutic optimum rather than simply the highest tested dose. The consistency of improvements across multiple physiological domains indicates that Naftalan oil's effects are both systemic and well-coordinated, supporting the hypothesis that its therapeutic mechanisms involve fundamental improvements in digestive function rather than simple symptomatic relief. The hematological improvements observed across all treatment groups provide important insights into Naftalan oil's broader physiological effects. The increases in erythrocyte count and hemoglobin levels suggest enhanced oxygen transport capacity, which could contribute to improved metabolic efficiency and overall animal health. The elevation in leukocyte counts, while remaining within normal ranges, indicates immune system stimulation that could enhance disease resistance without triggering inflammatory responses. The expert system's ability to integrate these diverse data streams and generate coherent treatment recommendations demonstrates the potential for artificial intelligence applications in veterinary medicine. The 95% accuracy rate achieved in dosage recommendations, combined with the 70% reduction in decision-making time, suggests that such systems could meaningfully enhance clinical practice efficiency while improving treatment outcomes. Perhaps most significantly, the adaptive learning capabilities of our system address one of the most challenging aspects of veterinary medicine: the need to continuously update and refine treatment protocols based on accumulating experience and evolving understanding. Traditional approaches to protocol development often lag years behind emerging research findings, while our system can incorporate new data and adjust recommendations in real-time. The implications of this work extend far beyond Naftalan oil therapy to encompass the broader integration of computational intelligence in veterinary practice. The methodological framework we have developed could be adapted to evaluate other natural therapeutic agents, analyze complex physiological interactions, and support clinical decision-making across diverse veterinary applications. The economic implications of improved therapeutic outcomes through expert system guidance could be substantial. Enhanced feed intake capacity directly translates to improved productivity and economic returns, while more precise dosing reduces waste and minimizes the risk of adverse effects. The time savings achieved through automated analysis could allow practitioners to manage larger case loads or devote more attention to complex cases requiring human expertise. From a broader perspective, our work contributes to the growing body of evidence supporting the integration of traditional therapeutic knowledge with modern analytical approaches. Rather than replacing clinical judgment, expert systems like ours can augment human decision-making by providing comprehensive data analysis and evidence-based recommendations that support rather than supplant veterinary expertise.

The development of our intelligent expert system for evaluating Naftalan oil effects on ovine gastrointestinal function represents a successful integration of empirical research, knowledge engineering, and clinical application. The clear demonstration of dose-response relationships across multiple physiological parameters provides a solid foundation for evidence-based treatment protocols, while the expert system architecture offers a scalable framework for similar applications across veterinary medicine. The identification of 20 ml as the optimal dosage for comprehensive therapeutic benefits provides practitioners with specific, evidence-based guidance for Naftalan oil therapy. The consistency of improvements across gastric pH, rumination time, feed intake capacity, and hematological parameters suggests that this dosage achieves true therapeutic optimization rather than simple parameter maximization. The expert system's demonstrated accuracy and efficiency in generating treatment recommendations validates the potential for artificial intelligence applications in veterinary practice. The combination of comprehensive data integration, sophisticated analytical capabilities, and adaptive learning ensures that the system remains current and continues to improve with accumulating experience. Future research directions should focus on expanding the knowledge base to encompass additional animal species, alternative dosing regimens, and combination therapies. The integration of real-time monitoring capabilities could enable dynamic treatment adjustment based on individual animal responses, further personalizing therapeutic approaches. Long-term studies evaluating the sustained effects of Naftalan oil therapy would provide valuable insights into optimal treatment duration and potential cumulative benefits. The methodology developed in this work offers a template for creating expert systems across diverse veterinary applications. The integration of experimental validation, knowledge engineering, and system design principles provides a comprehensive approach to developing decision support tools that can meaningfully enhance veterinary practice while maintaining the clinical judgment and expertise that remain central to animal healthcare. As veterinary medicine continues to evolve toward more precise, personalized, and evidence-based approaches, expert systems like the one we have developed will play increasingly important roles in supporting clinical decision-making, optimizing treatment outcomes, and advancing the scientific foundation of animal healthcare. The successful integration of traditional therapeutic knowledge with modern computational approaches demonstrated in this work provides a model for future developments in veterinary informatics and intelligent system applications.

1. Garg AX, Adhikari NK, McDonald H, Rosas-Arellano MP, Devereaux PJ, Beyene J, et al. Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: a systematic review. JAMA. 2005; 293: 1223-1238.
2. Maresova P, Penhaker M, Selamat A, Kuca K. The potential of artificial intelligence in healthcare. Neural Computing and Applications. 2018; 29: 555-563.
3. Güler ?, Y?ld?r?m S, Do?an, B. Artificial intelligence applications in animal husbandry: A review. Computers and Electronics in Agriculture. 2021; 188: 106353.
4. Turban E, Aronson, JE, Liang TP, Sharda R. Decision support and business intelligence systems (9th ed.). Pearson Education. 2011.
5. Jackson P. Introduction to expert systems (3rd ed.)Addison Wesley.1999.
6. Delen D. Predictive analytics: Data mining, machine learning and data science for practitioners. Pearson Education. 2020.
7. Ismayilov R, Abbasov E, Farzaliyev S. Naftalan oil and its applications in traditional and modern medicine. Azerbaijan J Med. 2017; 23: 17-22.
8. Mammadov I, Aliyev F. Use of Naftalan oil in veterinary medicine: Experimental studies. Azerbaijan Vet J. 2019; 6: 51-58.
9. Huseynov R, Mirzoyev A, Nagiyev T. Physiological effects of natural hydrocarbons in veterinary practice. Scientific Rep Azerbaijan Agriculture. 2020; 10: 91-98.
10. Durkin J. Expert systems: Design and development. Macmillan. 1994
11. Topal A, Duru, T. Veterinary expert systems and their development in livestock decision support. Turkish J Vet Res. 2021; 4: 39-45.
12. Nagiyev R, Abbasov S, Mahmudova, A. Pharmacodynamics of Naftalan hydrocarbons in livestock. J Exp Vet Med. 2021. 32: 110-117.
13. Kamilov Z, Taghiyev I, Musayev A. Dose-dependent effects of naphthalan oil in sheep. Vet Therapeutics. 2015; 13: 34-40.
14. Azizov A, Nasirov M., Rahimov T. Influence of Naftalan oil on biological markers in animal models. J Azerbaijani Med Res. 2020; 28: 63-70.
15. Guliyev R, Rasulov S. Therapeutic effects of naphthalan oil in gastroenterological disorders. Caspian J Med Sci. 2018; 4: 82-87.
16. Jain NC. Essentials of Veterinary Hematology. Lea Febiger. 1993.
17. Thrall MA, Weiser G, Allison RW, Campbell TW. Veterinary hematology and clinical chemistry(2nded.).Wiley-Blackwell. 2012.
18. Abdullayeva G, Mamedova S, Aliyev R. Veterinary hematology parameters in sheep under stress conditions. Vet Res J. 2022; 13: 45-51.
19. Berner ES. Clinical decision support systems: Theory and practice. Springer. 2007.
20. Mehmood M, Qamar AM, Nasir M. Veterinary diagnostics through intelligent decision support systems. Computer Methods and Programs in Biomedicine. 2022; 217: 106679.
21. Smith BP, Sherman DM. Large animal internal medicine (4th ed.). Elsevie. 2009.
22. Bhattacharya M, Mukherjee A, Chatterjee S. AI-based decision models in animal healthcare. Artificial Intelligence in Life Sciences. 2020; 1: 100003.
23. Radostits OM,Gay CC, Hinchcliff KW, Constable PD. Veterinary medicine: A textbook of the diseases of cattle, horses, sheep, pigs and goats (10th ed.). Elsevier Health Sciences. 2007.
24. Russom P. Big data analytics. TDWI Best Practices Report. 2011.