Leveraging artificial neural networks for optimizing Cinnamomum Sintoc essential oil production in Mount Ciremai

Aproveitando redes neurais artificiais para otimizar a produção de óleo essencial Cinnamomum Sintoc em Monte Ciremai

Abstract

Cinnamomum sintoc is a plant renowned for its production of high-quality essential oils. This study assessed the essential oil content in C. sintoc based on its morphological characteristics, environmental conditions, and soil nutrient composition. A mixed-methods approach was employed, with observational data collected from the Mount Ciremai National Park and analytical data obtained from soil laboratory tests. To analyze the data, Artificial Neural Networks (ANN), specifically ANFIS and MICMAC, were utilized for tasks such as pattern recognition, classification, and prediction. The results indicated that both environmental conditions and C. sintoc morphology have a direct influence on essential oil content, with soil nutrients identified as the most significant factor. The ANFIS analysis, with a prediction accuracy of 96%, identified the optimal nutrient levels for essential oil production: C/N ratio greater than 9, Mg at 0.5, C-organic above 6, N-total at 0.7, K at 0.6, P2O5 between 5 and 9, Ca below 5, and soil pH at 5.5. These findings suggest that the guidelines developed through ANFIS can be further refined into practical rules for optimizing the utilization of C. sintoc plants, potentially improving efficiency in essential oil production.

Keywords:
soil nutrients; Artificial Neural Network; ANFIS; MICMAC; sintoc morphology

Resumo

Cinnamomum sintoc é uma planta reconhecida por sua produção de óleos essenciais de alta qualidade. Este estudo avaliou o teor de óleo essencial em C. sintoc com base em suas características morfológicas, condições ambientais e composição de nutrientes do solo. Foi empregada uma abordagem de métodos mistos, com dados observacionais coletados do Parque Nacional de Monte Ciremai e dados analíticos obtidos por meio de testes laboratoriais de solo. Para analisar os dados, Redes Neurais Artificiais (RNA), especificamente ANFIS e MICMAC, foram utilizadas para tarefas como reconhecimento de padrões, classificação e previsão. Os resultados indicaram que tanto as condições ambientais quanto a morfologia de C. sintoc têm influência direta no conteúdo de óleo essencial, sendo os nutrientes do solo identificados como o fator mais significativo. A análise ANFIS, com precisão de previsão de 96%, identificou os níveis ideais de nutrientes para a produção de óleo essencial: razão C/N maior que 9, Mg a 0,5, C-orgânico acima de 6, N-total a 0,7, K a 0,6, P2O5 entre 5 e 9, Ca abaixo de 5 e pH do solo a 5,5. Essas descobertas sugerem que as directrizes desenvolvidas através da ANFIS podem ser ainda mais refinadas em regras práticas para otimizar o uso de plantas de C. sintoc, melhorando potencialmente a eficiência na produção de óleo essencial.

Palavras-chave:
nutrientes do solo; Rede Neural Artificial; ANFIS; MICMAC; morfologia do Sintoc

1. Introduction

Cinnamomum sintoc is a plant widely known for its aromatic properties and beneficial bioactive compounds (Chellappandian et al., 2018). Due to its antibacterial, anti-inflammatory, and aromatic qualities, these properties make it highly valuable in various industries, such as cosmetics and pharmaceuticals (Pandey et al., 2020). The importance of C. sintoc has been further emphasized in various studies (Ni et al., 2021; Ismail et al., 2021; Yulistyarini, 2020; Sharma et al., 2023; Giacometti et al., 2018; Chellappandian et al., 2018), which highlight its wide potential for application in various fields.

The production of essential oils from Cinnamomum sintoc is significantly influenced by various environmental factors, particularly the soil conditions in which the plant is cultivated (Ismail et al., 2024). Soil characteristics such as pH levels, moisture, and mineral and nutrient content play a pivotal role in determining the quality and quantity of essential oils produced (Dehsheikh et al., 2020). Therefore, a comprehensive understanding of the relationship between soil conditions and essential oil yields is essential for enhancing both the production and efficiency of C. sintoc cultivation.

Mount Ciremai, a protected area with a sensitive ecosystem, requires sustainable and efficient cultivation methods for C. sintoc to preserve the region’s ecological balance. Major challenge to this sustainability is the continued use of inefficient production techniques, where many producers rely on traditional technologies incapable of optimizing oil extraction. The distillation processes employed often fail to effectively separate the active components, leading to lower yields and inconsistent oil quality. Moreover, the environmental impact of essential oil production is concerning, as practices such as deforestation and pollution—resulting from the use of wood fuel in distillation—pose significant threats to the sustainability of Mount Ciremai’s ecosystem (Abbasi and Abbasi, 2010).

The quality of essential oils is fundamentally influenced by various factors, including production methods, cultivation conditions, and post-harvest handling practices (Sivakumar and Bautista-Baños, 2014). In Mount Ciremai, a natural habitat of Cinnamomum sintoc, optimizing these key determinants is crucial to achieving more efficient and higher-quality essential oil production.

To enhance both the quality and quantity of essential oils derived from C. sintoc, it is essential to focus on environmental stewardship, ensuring that the cultivation methods align with the ecological conditions of the plant's natural habitat (Hidayat et al., 2021). Research has demonstrated that the application of biostimulants can significantly improve plant height, chlorophyll content, relative moisture, biomass yield, and essential oil output compared to untreated control plants (Kapoor et al., 2004; Farruggia et al., 2024). More frequent applications of biostimulants further increase biomass and essential oil yield, indicating their potential in optimizing production.

The levels of essential oils produced by Cinnamomum sintoc (C. Sintoc) plants are greatly influenced by the fertility and chemical content of the soil where the plant grows. Soil fertility is directly related to the availability of nutrients needed by plants to grow optimally. C. Sintoc plants require certain nutrients such as nitrogen, phosphorus, and potassium which support vegetative growth and the formation of active chemical compounds in plants. A deficiency of one of these nutrients can inhibit the process of photosynthesis and plant metabolism, which ultimately results in a decrease in the levels of essential oils produced. moreover, the chemical content of the soil, such as soil pH and micronutrient levels, also affect the ability of plants to absorb nutrients and produce secondary compounds, including essential oils. Soil with an inappropriate pH, either too acidic or too alkaline, can inhibit the solubility and availability of nutrients needed by plants. This can cause an imbalance in plant metabolism and reduce the production of essential oils. Soil that has a high organic matter content is generally more fertile and supports overall plant growth, which in turn increases essential oil production. Good soil management, with proper fertilization according to nutrient needs, can improve soil fertility and increase the quality and quantity of essential oils produced by C. Sintoc. The control of soil quality, it will certainly also have implications for environmental sustainability so that the felling of plants in the natural environment around Mount Ciremai can be reduced with the efficiency of essential oil productivity from C. Sintoc.

In the developmental efforts of C. sintoc essential oil production in the Mount Ciremai region, it is imperative to make the production process efficient through technological advancements, including the selection of more productive C. sintoc trees. The maximum concentration of N, P, K, and Fe achieved the highest dry matter yield of peppermint (354.8 g/m2), essential oil content (2.7%), and essential oil yield (6.6 g/m2) (Ostadi et al., 2020). The nutrient levels contained in the soil K > N> Ca > Mg > P > S which are macro nutrients and Fe > Mn > Zn > B> Cu > Ni > Mo micro nutrients with a dose of 77 mg/dm3 have been proven to obtain maximum essential oil production, with adequate characteristics such as the chemical profile of lavender plants (Peçanha et al., 2021). Based on this, a model is needed to simulate the yield of essential oils based on the soil nutrient content in the area where C. sintoc grows.

One promising approach is the use of Artificial Neural Networks (ANN), a form of artificial intelligence capable of modeling complex relationships between variables that influence production yields (Rashid et al., 2021). Artificial Neural Networks (ANN) technology has been extensively applied across various fields, including agriculture and natural product processing, to predict and optimize production outputs. The application of Artificial Neural Networks (ANN) in the development of essential oils from C. sintoc can aid in forecasting factors affecting oil quality and yield and provide recommendations for more efficient and sustainable cultivation practices. Given the current challenges in the Mount Ciremai region, such as inconsistencies in yield quality, extraction process inefficiencies, and sustainable resource management, a more integrated and data-driven approach is essential for improving both the quality and quantity of essential oils produced.

In this study, Artificial Neural Networks (ANN) were employed to predict the development of essential oil yields from Cinnamomum sintoc by analyzing soil conditions in the Mount Ciremai region. Data related to soil properties—such as moisture content, pH, and nutrient levels—alongside the yield of essential oils were collected. Artificial Neural Network (ANN) models were then trained to identify correlations and patterns among these variables, to optimize essential oil production by predicting the most favorable soil conditions for C. sintoc cultivation.

The Artificial Neural Networks (ANN) model was designed to reveal the relationships between critical variables influencing production, allowing for the evaluation of its accuracy and efficiency in predicting essential oil yields. The model's performance will be compared to conventional methods in terms of predictive capacity and the optimization of the production process. This data-driven approach aims to provide actionable recommendations for improving the efficiency and quality of C. sintoc essential oil production in the Mount Ciremai area.

This research holds significant implications for the essential oil industry, particularly in the development of C. sintoc essential oils in Mount Ciremai. By leveraging Artificial Neural Networks (ANN) technology, it is anticipated that both production efficiency and quality will be enhanced. Furthermore, the findings of this study contribute to the advancement of data-based agricultural technologies and artificial intelligence in Indonesia, while offering economic and ecological benefits to local communities involved in essential oil production.

2. Research Methods

In this study, a mixed-method approach was used to combine field observation data with laboratory analysis results. This approach allows researchers to obtain a more comprehensive picture of the phenomenon being studied. Field observation data provides direct information from real conditions in the field, while laboratory data provides more controlled and accurate analysis results. The use of both types of data provides advantages in validating findings and increasing the accuracy of the analysis. For example, in a study that focuses on environmental factors that affect essential oil production, field data collection on weather conditions, soil, and plant quality can be combined with laboratory data on plant chemical composition and essential oil processing results (Creswell and Plano Clark, 2017).

The data sets obtained from both sources were then consolidated in tabular form to facilitate further analysis. This consolidation process is important to organize and align the data so that it can be analyzed more efficiently. Combining field and laboratory data in one tabular format ensures that the resulting information is easier to understand and can be directly used for predictive or analytical models. In this context, data tabulation makes it easier for researchers to identify the relationship between variables that influence the dependent variable, such as the effect of soil chemical nutrients on essential oil levels in C. Sintoc plants, which can then be used for the development of Artificial Neural Network (ANN)-based models (Tashman, 2019).

Next, the consolidated data is used as input for two types of artificial intelligence models, namely Adaptive Neuro-Fuzzy Inference System (ANFIS) and MICMAC artificial neural network. The ANFIS model, which combines the advantages of fuzzy logic and artificial neural networks, is very useful for handling uncertain and complex data, such as factors affecting essential oil production. Meanwhile, the MICMAC model, which focuses more on the analysis of relationships between variables, can provide a clearer picture of the relationship between factors affecting production results. Thus, these two models help researchers to formulate more accurate predictions and understand patterns that may not be visible in traditional analysis. The combination of mixed methods and these models enriches the research results and provides deeper insights into the processes being analyzed (Jang, 1993; Vester, 2007).

3. Data Collection

Data collection in this study was conducted through field observations and laboratory analyses. Field observations were initiated using a purposive sampling method, whereby Simplicia from the leaves, bark, and soil scum were collected upon first encountering the Cinnamomum sintoc tree, based on its diameter class. Environmental parameters, such as temperature and humidity, which could influence the morphology of C. sintoc, were also recorded. Morphological data were used as a reference to assess their impact on essential oil yields. According to research by Ismail et al. (2023), plant morphology and environmental conditions significantly influence the yield of essential oils derived from C. sintoc plants. Field observation data on the environmental conditions of C. sintoc growth were collected from sites around Mount Ciremai. The observation location map is presented in Figure 1 below.

Figure 1
Map of the location of soil sampling and C. Sintok plant Simplicia on Mount Ciremai.

Environmental data, including temperature, humidity, light intensity, and rainfall, were collected from the observation sites using environmental meters placed in each sampling plot. These sampling plots were established to gather plant morphology data, such as plant height, stem diameter, and growth stages. In addition, soil samples were collected from each plot for laboratory analysis, yielding information on soil chemistry, including pH levels, nitrogen, phosphorus, potassium, and other essential micronutrients.

4. Data Preparation

4.1. Environmental data collection

Environmental data from the Cinnamomum sintoc growth sites were collected through continuous observations and direct measurements of key environmental parameters, including temperature, humidity, light intensity, and rainfall. Figure 2 below illustrates the environmental data collection process.

Figure 2
Soil sampling process and plant simplicia C. Sintok.

4.2. Soil chemical data collection

Soil chemical data were obtained through laboratory analysis of soil samples systematically and representatively collected from the research site to ensure that the data accurately reflected field conditions. In the laboratory shown in Figure 3 several key soil chemical parameters were analyzed, including pH, organic matter content, cation exchange capacity (CEC), and essential nutrients such as nitrogen (N), phosphorus (P), and potassium (K). These analyses followed scientifically recognized standard methods to ensure the accuracy and reliability of the results. The data were then used to assess the soil fertility characteristics at the sites where Cinnamomum sintoc trees grow. In addition, representative leaf and bark samples from C. sintoc, which were used for essential oil distillation, were collected alongside the soil samples. Previous research has shown that the use of combined fertilizers can enhance leaf dry matter, improve plant nutritional value, and increase essential oil content (Yousefzadeh et al., 2015; Mavandi et al., 2021).

Figure 3
The process of analyzing soil chemical elements in the laboratory .

4.3. Data collection of essential oils

Essential oils are extracted from the leaves and bark of Cinnamomum sintoc through a distillation process. Leaf and bark samples were collected based on the sampling plots illustrated in Figure 4 and from the same locations where soil samples were taken for chemical analysis. This ensures that the essential oil distillation results are influenced by the corresponding environmental and soil factors. The distillation process is depicted in Figure 4 below. Essential oils were sampled from multiple distillate batches to capture diverse results. To minimize variability, it is essential to ensure that samples are collected under consistent conditions and at regular intervals. The total volume of essential oils obtained from each batch was also measured and recorded to assess production efficiency.

Figure 4
Essential oil distillation Process.

4.4. Data analysis framework utilizing Artificial Neural Networks

Artificial Neural Networks (ARTIFICIAL NEURAL NETWORKS (ANN)s) are computational systems inspired by biological neural networks, designed to model complex relationships between inputs and outputs or to identify patterns within data. In data analysis, ARTIFICIAL NEURAL NETWORKS (ANN)s are utilized for tasks such as pattern recognition, classification, and prediction. This study employs two specific models: ANFIS and MICMAC.

The Adaptive Neuro-Fuzzy Inference System (ANFIS) is a hybrid model that integrates neural networks with fuzzy logic to create a flexible and adaptive learning system (Jang, 1993). ANFIS is capable of learning from data adaptively by optimizing existing fuzzy rules. Its interpretative abilities, combined with fuzzy logic, enable the system to generate outputs that can be understood as predictive logic (Vargas et al., 2024).

The MICMAC method, on the other hand, is a structural analysis tool used to explore and map relationships between variables in complex systems (Zhao et al., 2018). MICMAC helps identify variables with significant influence and dependency within the system, providing a structured method for determining the ranking of variables or elements. This ranking allows for the development of inter-variable linkages and strategic plartificial neural networks (ANN)ing in a more organized martificial neural networks (ANN)er.

By using MICMAC, the influence of each variable on the system can be systematically evaluated. The results are then integrated with the ANFIS model to predict how specific variables will affect the system’s output. This combination allows for a robust prediction model, where MICMAC provides a clear understanding of the variable interactions, and ANFIS outputs predictions regarding the influence of these variables on the desired outcomes.

4.5. Adaptive Neuro-Fuzzy Inference System (ANFIS)

In this study, ANFIS serves as the core inference system, merging neural networks with fuzzy logic to adapt to the data and generate accurate predictions. The initial design of ANFIS will guide the development of the architecture, ensuring it meets the requirements for effective data modeling and prediction.

In this study, the ANFIS method consists of five primary layers. The first layer, known as fuzzification, marks the initial stage of the ANFIS process. At this stage, the input data, which includes the chemical properties of the Cinnamomum sintoc growth location and essential oil yield predictions, is processed. The subsequent layers include layer 2, responsible for fuzzy logic operations, layer 3, which handles the normalization of the activation degree, layer 4, dedicated to defuzzification, and finally, layer 5, where the system outputs its predictions.

Additionally, a hybrid algorithm is employed within the ANFIS model to optimize its parameters. This hybrid approach combines two learning processes: forward learning and backward learning. In forward learning, the Least Squares Estimation (LSE) recursive algorithm is used, while in backward learning, the Error Backpropagation (EBP) algorithm is applied. The accuracy of the predictions is evaluated by calculating the root mean square error (RMSE) to measure the prediction error rate. The methodological framework of this research is depicted in Figure 5.

Figure 5
ANFIS architectural design for essential oil prediction.

Figure 6 illustrates two types of nodes: adaptive nodes, represented by square symbols, and fixed nodes, represented by circular symbols. The output from each layer is denoted as \( O^i_j \), where \( i \) corresponds to the layer number and \( j \) represents the rule number.

Figure 6
ANFIS System Development Design.

Kusumadewi and Hartati (2010) outline the ANFIS network as comprising five distinct layers, each performing a specific function in the network's overall structure and operation.

4.6. Layer 1 (fuzzy layer)

The input layer serves as a receiver of raw data from various sources. In this context, the relevant inputs are:

• Soil PH: Measures the acidity or alkalinity level of the soil.

• Nitrogen Content (N): Indicates the amount of nitrogen in the soil.

• Phosphorus Content (P): Measures the amount of phosphorus available.

• Potassium Content (K): Indicates the amount of potassium available.

• Other Organic Content: May include elements such as Magnesium (Mg), Calcium (Ca), and other organic ingredients

The first layer shows Every node in this layer is adaptive with a node function and how much each numerical input corresponds to a different fuzzy set. MF from fuzzy set Ai becomes Bi, then the output of this layer can be calculated by the Formula 1-2.

$$Oi=\mu Ai\left(x\right);i$$ (1)

$$Oi=\mu Bi\left(y\right);i$$ (2)

where x is the input to node i, Ai is the linguistic variable associated with this node and $\text{μ}$ Ai is the membership function of Ai. Usually $\text{μ}$_Ai (x) is chosen as where x is the input and {ai, bi, ci} is the premise parameter set seen in the Formula 3-4.

${\text{μ}}_{\text{A}}\left({\text{x}}_{\text{t}}\right)={\text{e}}^{\frac{-{({\text{x}}_{\text{t}}-\text{c})}^2}{2{\text{σ}}^2}}$ (3)

$\text{f x};{\text{a}}_{\text{i}},{\text{b}}_{\text{i}},{\text{c}}_{\text{i}}=\frac{1}{1+\frac{\text{x}-{\text{c}}_{\text{i}}{}^{2{\text{b}}_{\text{i}}}}{{\text{a}}_{\text{i}}}}$ (4)

4.7. Layer 2 (product layer)

Each node that is in layer 2 is a non-adaptive node, which means that the parameter value is fixed. Here, the input data is translated into a fuzzy form. Each input is mapped to a fuzzy membership function. For example:

• Soil pH: "Acidic", "Neutral", "Alkaline"

• Nitrogen Content: "Low", "Medium", "High"

The function of this node multiplies each incoming input signal as seen in Formula 5 as follows:

${\text{O}}_{\mathrm{2,}\text{i}}={\text{w}}_{\text{i}}={\text{μA}}_{\text{i}}\left(\text{x}\right).{\text{μB}}_{\text{i}}\left(\text{y}\right);\text{i}=\mathrm{1,2}$ (5)

Each output node expresses the degree of activation on each fuzzy rule. The number of rules formed follows the number of nodes in this layer.

4.8. Layer 3 (normalization layer)

Each node in this layer is a non-adaptive node representing a normalized degree function, expressed as the output ratio. This layer applies rules grounded in fuzzy logic. Each rule follows an IF-THEN structure, such as:

• Rule 1: IF pH Soil is "Exceptional" AND Nitrogen is "High" THEN Essential Oil is "High"

• Rule 2: IF Phosphorus is "Low" THEN Essential Oil is "Low"

The function of node I in the previous layer is as follows in Formula 6:

${\text{O}}_{\mathrm{3,1}}={\text{w}}_{\text{t}}\frac{{\text{w}}_{\text{i}}}{{\text{w}}_1+{\text{w}}_2},\text{dengan i}=\mathrm{1,2}$ (6)

If there are more than 2 rules, then the function can be expanded by dividing by the total number of watts for all rules.

4.9. Layer 4 (defuzzification layer)

Each node in this layer is an adaptive node with the following node function in Formula 7-8:

${\text{O}}_{\mathrm{4,1}\text{t}}={\text{w}}_{1\text{t}}^{\text{*}}{\text{Z}}_{\text{t}}^{\left(1\right)}={\text{w}}_{1\text{t}}^{\text{*}}\left({\text{α}}_1{\text{Z}}_{\text{t}-1}+{\text{β}}_1{\text{Z}}_{\text{t}-2}+{\text{γ}}_1\right)$ (7)

${\text{O}}_{\mathrm{4,2}\text{t}}={\text{w}}_{2\text{t}}^{\text{*}}{\text{Z}}_{\text{t}}^{\left(2\right)}={\text{w}}_2^{\text{*}}\left({\text{α}}_2{\text{Z}}_{\text{t}-1}+{\text{β}}_2{\text{Z}}_{\text{t}-2}+{\text{γ}}_2\right)$ (8)

Where, αi, βi, γi are the set of parameters of the node and are called consequent parameters.

4.10. Total Output Layer (layer 5)

This is the last layer that summarizes all the inputs with the following node function in Formula 9:

${\text{O}}_{5\text{t}}={\hat{\text{Z}}}_{\text{t}}={\text{w}}_{1\text{t}}^{\text{*}}{\text{Z}}_{\text{t}}^{\left(1\right)}+{\text{w}}_{2\text{t}}^{\text{*}}{\text{Z}}_{\text{t}}^{\left(2\right)}$ (9)

4.11. MicMac analysis

The MICMAC (Matrix of Crossed Impact Multiplications Applied to a Classification) method is a structural analysis technique introduced by Duperrin and Godet (1973). It offers a systematic approach to addressing complexity by ranking the elements of a system based on the relationships between variables. MICMAC is commonly used to identify key factors within a system (Barati et al, 2019). One of its strengths is its ability to transform qualitative data into quantitative insights by applying matrix properties (Sinkovics and Alfoldi, 2012). The capacity to group and prioritize strategic variables and their interdependencies provides a more reliable foundation for addressing complex issues. Using MICMAC, critical system variables can be effectively identified and analyzed (Ariyani & Fauzi, 2019).

The MICMAC method employs an analytical framework based on two axes: driver power (DP) and dependency (D) (Zhao et al., 2024). This allows variables to be categorized into sectors, clusters, or quadrants (D’Souza et al., 2007). The method involves three fundamental steps: identifying relevant variables, explaining the relationships between them, and pinpointing key variables (Putra and Pramesti, 2019). In the context of this study, which addresses the challenges related to essential oil production based on the environmental conditions and morphology of Cinnamomum sintoc, the objective is to assess the influence of these factors on essential oil yields. The analysis of questionnaire data using MICMAC involves converting the weight of each variable into a matrix of direct influence (MDI), as shown in Figure 7.

Figure 7
The framework of MICMAC Analisis

MICMAC analysis consists of two main stages. The first stage focuses on understanding the scope of the problem and the system under investigation (Abbas et al., 2022). The overall flow of MICMAC analysis is illustrated in Figure 8.

Figure 8
Illustration of MICMAC Analysis Results

Based on the results of field observations, several key variables have been identified and quantified concerning the constructed variables, resulting in a matrix of direct influence, as shown in Table 1.

Table 1
Identification of dimensions and variable influence on essential oil results.

N"	Long Label	Short Label	Theme
1	Effect of C. Sintoc tree diameter on essential oil yield	EDtSY	Morphology
2	Effect of C. sintoc tree height on essential oil yield	EHtSY	Morphology
3	The height of the growing site relative to the yield of essential oils	MaslY	Environmental
4	Growing environmental conditions on essential oil yields	TESg	Environmental
5	The direction of the slope where C. sintoc grows	ScY	Environmental
6	Soil chemicals for the influence of essential oil yields	SdS	Environmental
7	Humidity of the growing location	HmdES	Environmental

Using the MICMAC method, Figure 9 presents the Matrix of Direct Influence (MDI), which has been repositioned into a variable map. This map visually represents the influence-dependence relationship of the variables, categorizing them into four distinct sectors or quadrants. These quadrants provide a clearer understanding of the positioning and interdependencies among the variables.

Figure 9
Position of a system variable in the direct influence-dependence map.

The data collected in this study consists of both primary and secondary sources. Seven key dimensions were identified as influencing essential oil yield: tree diameter, tree height, altitude of the growth location, environmental temperature, soil chemical composition, slope direction, and humidity. These dimensions serve as the foundation for constructing the attributes that affect essential oil production.

The variables used in the analysis were derived from direct field observations, laboratory analyses, and detailed measurements. Data collection was conducted around Mount Ciremai, with additional analyses performed in the Soil Chemistry Laboratory. The relationships between the variables were quantified using a scale from 0 to 3, with an additional category, P, following the methodology of Godet (2000): 0 = no relationship (non-existent); 1 = weak relationship (low direct influence); 2 = moderate relationship (medium direct influence); 3 = strong relationship (high direct influence); and P = potential influence

This quantification approach enables a structured analysis of the direct relationships between variables, providing a more comprehensive understanding of their impact on essential oil yield.

4.12. Linkages and synergies between ANFIS and MICMAC

The integration of ANFIS and MICMAC in data analysis offers distinct advantages, particularly for understanding complex and dynamic systems. ANFIS effectively addresses non-linear systems by combining adaptive learning and fuzzy logic, making it highly suited for environments characterized by uncertainty and variability. MICMAC, on the other hand, analyzes the structural relationships and dynamics within the system, allowing for the identification of key factors that significantly influence the system’s behavior. Together, they enhance decision-making, with ANFIS providing accurate predictions based on historical data and patterns, and MICMAC offering strategic guidance by identifying critical variables and their interdependencies, facilitating scenario artificial neural networks (ANN) planning and risk mitigation.

This combination has practical applications across various fields. In industry, ANFIS can be used for demand forecasting, while MICMAC optimizes supply chain analysis, streamlining production processes. In economics, ANFIS helps predict economic trends, and MICMAC contributes to macroeconomic policy analysis and decision-making. In environmental studies, ANFIS models the impacts of climate change, and MICMAC explores the interrelationships between environmental factors. By leveraging the strengths of both methods, this approach provides a comprehensive understanding of complex systems and supports more informed, strategic decision-making across multiple disciplines.

5. Discussion and Results

5.1. Matrix of direct influence

5.1.1. Direct Influence Matrix (Matrix of Direct Influence)

Based on the analysis of field data, the essential oil yield of C. sintoc in its growing environment around Mount Ciremai was examined in relation to various environmental factors. The input data comprised seven key variables: tree diameter, tree height, altitude, temperature of the growing environment, soil chemical composition, slope direction, and humidity. These seven variables, as outlined in Table 2, were grouped accordingly for analysis. Each variable was subsequently incorporated into the MICMAC analysis, where they were evaluated using a matrix of direct influences (MDI). The analysis of these seven variables, which are considered crucial for determining essential oil levels, demonstrated consistency and stability in the results. The MDI matrix is provided in Table 2 below.

Table 2
Direct relationship of each variable to essential oil content.

The relationship of each variable
	EDtSY	EHtSY	MaslY	TESg	ScY	SdS	HmdES	EOVol
EDtSY	0	3	2	2	2	2	1	3
EHtSY	3	0	2	2	2	2	1	3
MaslY	1	1	0	1	2	1	2	2
TESg	3	2	1	0	2	2	2	3
ScY	2	2	1	1	0	2	1	1
SdS	3	3	1	1	1	0	2	3
HmdES	1	1	1	3	1	1	0	1
EOVol	3	3	2	3	1	3	1	0

Key variables influencing the essential oil content of C. sintoc around Mount Ciremai, considering both the position and intensity of their direct or indirect effects (and the absence of causal relationships), have demonstrated the validity and robustness of this approach in identifying the most significant variable: the growing environmental conditions. These conditions are identified as the key factor, serving as a pivotal variable for predicting future outcomes and benefits. An essential aspect of this analysis is identifying key environmental factors, which play a critical role in determining essential oil yield. The power relationships between these variables are visually represented in the influence-dependence quadrant, as shown in Figure 10.

Figure 10
Graphic Illustration of Influence-Dependence Variables Matrix.

Based on the results of the MICMAC analysis presented in Figure 10, the variables identified in the first quadrant (determinant variables) include the environmental conditions of the growth location (TESg) and the soil chemical composition influencing essential oil yields (SdS). The key characteristic of these variables is their high degree of influence combined with low dependence.

In the second quadrant (key variables), which includes variables with both high influence and high dependence but unstable relationships, the primary factors are the effect of Cinnamomum sintoc tree height on essential oil yield (EHtSY) and the impact of tree diameter on essential oil yield (EDtSY).

In the third quadrant (result variables), characterized by low influence but high dependence, the relevant variables include the slope direction of the C. sintoc growth site (ScY), altitude (MaslY), and the humidity of the growing location (HmdES). The fourth quadrant (autonomous variables) typically contains variables with low influence and low dependence. However, in this analysis, no variables exhibited very low influence on essential oil content, as all the data relate to critical environmental factors.

The relationships between these variables are further illustrated in Figure 10. In this graph, the connections are color-coded: green lines represent weak influences, blue lines indicate moderate influences, dark blue lines show relatively strong influences and red lines denote the strongest influences.

From Figure 10, it is evident that the most influential variables in determining essential oil levels, based on the growth conditions of C. sintoc around Mount Ciremai, are environmental conditions (TESg), soil chemicals affecting essential oil yields (SdS), tree height (EHtSY), and tree diameter (EDtSY). These variables exert the strongest influence on essential oil production. In contrast, variables with the lowest influence include the slope direction of the growth location (ScY), altitude (MaslY), and humidity (HmdES).

In conclusion, the determinant variables—environmental conditions, soil chemical composition, tree height, and tree diameter—play a critical role in influencing essential oil levels in C. sintoc around Mount Ciremai. Enhancing these factors can directly contribute to increasing essential oil yields.

5.1.2. Matrix of Indirect Influence

In addition to the Matrix of Direct Influence (MDI), the position of variables in the influence-dependence quadrant is also assessed using the Matrix of Indirect Influence (MII). This allows for the identification of positional shifts in the variables, as illustrated by the displacement map. Based on the MII analysis, each system variable is reclassified into four quadrants according to their position in the influence-dependence chart, as shown in Figure 11. The analysis reveals that certain variables, such as environmental conditions of the growth location (TESg) and soil chemical composition affecting essential oil yields (SdS), experience a change in position. This shift underscores the critical role these variables play, highlighting them as the most significant factors influencing the essential oil levels of Cinnamomum sintoc growing around Mount Ciremai.

Figure 11
Matrix of Indirect Influence.

Visually, the percentage of the complexity of interactions between system variables related to the level of influence and its indirect dependence (indirect) influence on other variables is shown in Figure 12 below. The variables that are the main determinants of essential oil levels are the environmental conditions of the growing location (TESg), and soil chemicals for the influence of essential oil yields (SdS). In addition, it can also be seen between the plant height variable and the plant diameter variable as well as other variables that are blue arrows indirectly becoming one of the main determinants of essential oil levels. The number on each arrow indicates the magnitude of the degree or rating of influence obtained through the iteration of the Boolean matrix. In contrast to direct influence, most variables have a very strong dependence on other variables (marked by many red lines).

Figure 12
Potential Indirect Influence Graph.

5.1.3. NeuroFuzzy-ANFIS implementation

In this study, the design was develope using primary data from the collective results of data from laboratory observations and analysis. Based on the results of the analysis using MicMac, the chemical content of the soil is one of the absolute factors in influencing the volume of essential oils obtained from sintoc (Cinnamomum sintoc) that grows around Mount Ciemai. So to sharpen the discussion and continue the results of the MicMac analysis, it was continued using ANFIS Artificial Neural Network analysis. Data input is the result of observation and observation from soil laboratories as input variables, namely C Organic, N Total, P2O5, C/N, Ca, Mg, Ph to ANFIS, then ANFIS automatically builds a data input model that forms rules as a reference to make predictions (Figure 13).

Figure 13
Data input and model construction on Neuro Fuzzy- ANFIS.

The ANFIS model evaluates the accuracy of its training process through the ANFIS Editor graph, where training data are represented by blue circles, and the results of the ANFIS training are denoted by red stars, as illustrated in Figure 14a. The closer the red stars align with the blue circles, the more accurate the ANFIS model is in training the data, resulting in a smaller average training error. A lower error value, approaching the defined error tolerance, indicates a well-designed system with high accuracy in generating output data. Conversely, if the red stars deviate significantly from the blue circles, it reflects poorer ANFIS performance in testing the data.

Figure 14
ANFIS programming input data training.

A new system can be deemed successful and ready for application if the test results show a low average error, approaching the target error tolerance of 0. In Figure 14b, the close alignment of the red stars within the blue circles suggests that the input data exhibits an error value close to 0. Therefore, it can be concluded that the input variables—trust, commitment, communication, and satisfaction—effectively represent a reliable model for predicting essential oil content based on soil chemistry, as processed by the ANFIS program.

Based on the input data into the ANFIS program, ANFIS programming automatically builds its own rules from the data. From the input data of 8 variables to see the volume of essential oils in a total of 6561 rules formed by ANFIS (Figure 15a), if the parameters in the rule viewer are moved manually, the changes in the model will be seen directly. Of the 6561 rules formed by ANFIS, one example of rules is taken, there is in Figure 15b develope by ANFIS in the assessment to be one of the best results.

Figure 15
(a) Rules Develope by ANFIS programming, (b) ANFIS Rule viewer.

Figure 16 presents the 3D surface visualizations generated by the ANFIS model. In Figure 16a, the relationship between the C-Organic content and N_Total in the soil and essential oil yield is depicted. The maximum yield is achieved when the C-Organic content reaches 10 and N_Total is at its peak value, level 1.

Figure 16
Illustrates the surface visualizations from the ANFIS model results: (a) the effect of soil chemical variables C_Organik and N_total on essential oil volume, (b) the influence of Ca and Mg on essential oil volume, and (c) the impact of K and P2O5 on essential oil volume.

Figure 16b illustrates the influence of soil chemical components, specifically calcium (Ca) and magnesium (Mg), on essential oil yield. The 3D surface shows that higher Ca levels significantly enhance the yield, indicating that Ca plays a crucial role in optimizing essential oil production. Similarly, Mg levels within the range of 0.6 to 0.8 are shown to optimize essential oil volume, with maximum yields observed when Ca levels range between 5 and 10.

Figure 16c demonstrates the effect of potassium (K) and phosphorus (P2O5) on essential oil production. The optimal K levels for maximizing essential oil volume are between 0.4 and 0.8, while P2O5 levels between 6 and 8 are most favorable for achieving maximum yield. These visualizations generated by ANFIS provide valuable guidelines for optimizing essential oil production by indicating the ideal soil chemical conditions for maximizing yield efficiency.

The relationship between soil chemical composition and essential oil yield is analyzed by visualizing variables such as C/N ratio, Mg, C-organic, N-total, K, P2O5, Ca, and pH in relation to essential oil volume. Soil characteristics, including pH, moisture, and mineral composition, significantly influence the plant’s ability to produce essential oils. For instance, an improper soil pH can disrupt the plant's metabolic processes, leading to a reduction in essential oil production. In contrast, soils enriched with essential nutrients such as nitrogen and phosphorus support better plant growth, enhancing essential oil content.

Additionally, organic matter in the soil plays a critical role in essential oil production. Soils with high organic content have improved water retention and provide a stable environment for root growth, promoting plant health and increasing essential oil yields. The addition of organic matter, such as compost or green manure, can further improve nutrient availability and soil structure, positively impacting essential oil production.

Moreover, the presence of micronutrients like magnesium, calcium, and potassium also influences both the quality and quantity of essential oils. Imbalances in these microelement levels can induce stress in plants, impairing the synthesis of essential oils. Therefore, regular monitoring and management of soil chemical content are essential for optimizing essential oil production. Further research on these interactions could deepen our understanding of how proper soil management enhances essential oil yields. The Figure 17 below illustrates the influence of various macro- and micronutrients on the volume of essential oils produced.

Figure 17
The Effect of Contained Soil Chemical Elements on the Prediction of Essential Oil Content.

5.1.4. Prediction accuracy

In this study, the predictive accuracy of the Adaptive Neuro-Fuzzy Inference System (ANFIS) reached an impressive 96%. This high level of precision is clearly illustrated in Figure 18, which compares the actual data from essential oil distillation with the predicted data generated by the ANFIS model. The graph demonstrates that the model’s predictions closely align with the actual data, indicating that ANFIS effectively captures and maps the complex relationships among the variables involved.

Figure 18
Accuracy of Essential Oil Content Prediction based on analysis data and observation data.

The analysis confirms that soil chemical variables significantly impact the volume of essential oils produced. The key variables examined include the C/N ratio, magnesium (Mg) content, organic carbon (C-organic), total nitrogen (N-total), potassium (K), phosphate (P2O5), calcium (Ca), and soil pH. The visualization of the results shows that each variable contributes to essential oil yield, with varying degrees of influence.

Overall, the findings of this study highlight the crucial role of soil chemistry in maximizing essential oil production. The ANFIS analysis reveals a strong relationship between soil conditions and essential oil yield, suggesting that optimal management of these factors can significantly enhance production. These results provide a solid foundation for developing improved land management strategies aimed at increasing both the quantity and quality of essential oils in the future.

6. Conclusion

The results of the MICMAC analysis indicate that both environmental conditions and the morphology of Cinnamomum sintoc have a direct impact on essential oil content. Among all the variables examined, soil nutrients emerge as the most significant factor influencing essential oil production. According to the ANFIS analysis, which achieved a prediction accuracy of 96%, the optimal levels of key soil nutrients for maximizing essential oil yields are as follows: a C/N ratio greater than 9, magnesium (Mg) at 0.5, organic carbon (C-organic) above 6, total nitrogen (N-total) at 0.7, potassium (K) at 0.6, phosphate (P2O5) between 5 and 9, calcium (Ca) below 5, and soil pH at 5.5.

These findings suggest that the nutrient thresholds identified by ANFIS can be incorporated into rules for developing IoT-based smart farming applications. Such systems could automate fertilization and irrigation for C. sintoc cultivation, optimizing essential oil production through precise, data-driven management.

Further research could focus on refining these IoT applications by integrating real-time soil monitoring systems, which would adjust nutrient inputs dynamically based on evolving environmental conditions. Additionally, future studies could explore the long-term effects of these optimized nutrient levels on essential oil quality and sustainability, potentially expanding the model to include other plant species and regions. This research would contribute to the advancement of precision agriculture, enhancing both yield and resource efficiency.

References

• ABBAS, H., ASIM, Z., AHMED, Z. and MOOSA, S., 2022. Exploring and establishing the barriers to sustainable humanitarian supply chains using fuzzy interpretive structural modeling and fuzzy MICMAC analysis. Social Responsibility Journal, vol. 18, no. 8, pp. 1463-1484. http://doi.org/10.1108/SRJ-12-2020-0485
  » http://doi.org/10.1108/SRJ-12-2020-0485
• ABBASI, T. and ABBASI, S.A., 2010. Biomass energy and the environmental impacts associated with its production and utilization. Renewable & Sustainable Energy Reviews, vol. 14, no. 3, pp. 919-937. http://doi.org/10.1016/j.rser.2009.11.006
  » http://doi.org/10.1016/j.rser.2009.11.006
• ARIYANI, N. and FAUZI, A., 2019. Analysis of strategic variables for ecotourism development; an application of Micmac. South Asian Journal of Social Studies and Economics, vol. 3, no. 3, pp. 1-12. http://doi.org/10.9734/sajsse/2019/v3i330107
  » http://doi.org/10.9734/sajsse/2019/v3i330107
• BARATI, A.A., AZADI, H., DEHGHANI POUR, M., LEBAILLY, P. and QAFORI, M., 2019. Determining key agricultural strategic factors using AHP-MICMAC. Sustainability (Basel), vol. 11, no. 14, pp. 3947. http://doi.org/10.3390/su11143947
  » http://doi.org/10.3390/su11143947
• CHELLAPPANDIAN, M., VASANTHA-SRINIVASAN, P., SENTHIL-NATHAN, S., KARTHI, S., THANIGAIVEL, A., PONSANKAR, A., KALAIVANI, K. and HUNTER, W.B., 2018. Botanical essential oils and uses as mosquitocides and repellents against dengue. Environment International, vol. 113, pp. 214-230. http://doi.org/10.1016/j.envint.2017.12.038 PMid:29453089.
  » http://doi.org/10.1016/j.envint.2017.12.038
• CRESWELL, J.W. and PLANO CLARK, V.L., 2017. Designing and conducting mixed methods research 3rd ed. Thousand Oaks: Sage Publications.
• D’SOUZA, C., TAGHIAN, M. and KHOSLA, R., 2007. Examination of environmental beliefs and its impact on the influence of price, quality and demographic characteristics with respect to green purchase intention. J Target Meas Anal Mark, vol. 15, no. 2, pp. 69-78. http://doi.org/10.1057/palgrave.jt.5750039
  » http://doi.org/10.1057/palgrave.jt.5750039
• DEHSHEIKH, A.B., SOURESTANI, M.M., ZOLFAGHARI, M. and ENAYATIZAMIR, N., 2020. Changes in soil microbial activity, essential oil quantity, and quality of Thai basil as response to biofertilizers and humic acid. Journal of Cleaner Production, vol. 256, pp. 120439. http://doi.org/10.1016/j.jclepro.2020.120439
  » http://doi.org/10.1016/j.jclepro.2020.120439
• DUPERRIN, J.C. and GODET, M., 1973. Methode de hierarchisation des elements d’un systeme. Rapport Economique du CEA, pp. 49–51.
• FARRUGGIA, D., DI MICELI, G., LICATA, M., LETO, C., SALAMONE, F. and NOVAK, J., 2024. Foliar application of various biostimulants produces contrasting response on yield, essential oil and chemical properties of organically grown sage (Salvia officinalis L.). Frontiers in Plant Science, vol. 15, pp. 1397489. http://doi.org/10.3389/fpls.2024.1397489 PMid:39011298.
  » http://doi.org/10.3389/fpls.2024.1397489
• GIACOMETTI, J., BURSAĆ KOVAČEVIĆ, D., PUTNIK, P., GABRIĆ, D., BILUŠIĆ, T., KREŠIĆ, G., STULIĆ, V., BARBA, F.J., CHEMAT, F., BARBOSA-CÁNOVAS, G. and REŽEK JAMBRAK, A., 2018. Extraction of bioactive compounds and essential oils from mediterranean herbs by conventional and green innovative techniques: a review. Food Research International, vol. 113, pp. 245-262. http://doi.org/10.1016/j.foodres.2018.06.036 PMid:30195519.
  » http://doi.org/10.1016/j.foodres.2018.06.036
• GODET, M., 2000. The art of scenarios and strategic planning: tools and pitfalls. Technological Forecasting and Social Change, vol. 65, no. 1, pp. 3-22. http://doi.org/10.1016/S0040-1625(99)00120-1
  » http://doi.org/10.1016/S0040-1625(99)00120-1
• HIDAYAT, S., ZUHUD, E.A.M., WIDYATMOKO, D. and BAHRUNI, B., 2021. Assessing the conservation value of medicinal plant collections in Bogor Botanic Gardens, Indonesia. Biodiversitas Journal of Biological Diversity, vol. 22, no. 2. http://doi.org/10.13057/biodiv/d220221
  » http://doi.org/10.13057/biodiv/d220221
• ISMAIL, A., HENDRAYANA, Y., RAMADANI, D. and UMIYATI, S., 2021. Composition of vegetation types and structures in gunung Ciremai National Park Forest. IOP Conference Series. Earth and Environmental Science, vol. 748, no. 1, pp. 012009. http://doi.org/10.1088/1755-1315/748/1/012009
  » http://doi.org/10.1088/1755-1315/748/1/012009
• ISMAIL, A.Y., KUSMANA, C., NAINGGOLAN, M.F. and GANDASECA, S., 2024. Prediction of essential oil content in C. sintoc Leaves based on the direction of vegetation slope in Mount Ciremai National Park using ANFIS Artificial Neural Network. Brazilian Journal of Biology = Revista Brasileira de Biologia, vol. 84, pp. e282113. http://doi.org/10.1590/1519-6984.282113 PMid:38808789.
  » http://doi.org/10.1590/1519-6984.282113
• ISMAIL, A.Y., KUSMANA, C., SUDIANA, E., AYU ANDAYANI, S., AMINUDIN, S. and FERNANDO NAINGGOLAN, M., 2023. Exploring and assessing Cinnamomum sintoc Blume essential oil composition in lowland forests of Mount Ciremai National Park, West Java. Forestry Ideas, vol. 29, no. 2(66), pp. 327-339.
• JANG, J.-S.R., 1993. ANFIS: adaptive-network-based fuzzy inference system. IEEE Transactions on Systems, Man, and Cybernetics, vol. 23, no. 3, pp. 665-685. http://doi.org/10.1109/21.256541
  » http://doi.org/10.1109/21.256541
• KAPOOR, R., GIRI, B. and MUKERJI, K.G., 2004. Improved growth and essential oil yield and quality in Foeniculum vulgare mill on mycorrhizal inoculation supplemented with P-fertilizer. Bioresource Technology, vol. 93, no. 3, pp. 307-311. http://doi.org/10.1016/j.biortech.2003.10.028 PMid:15062827.
  » http://doi.org/10.1016/j.biortech.2003.10.028
• KUSUMADEWI, S. and HARTATI, S., 2010. Neuro-Fuzzy Integration of Fuzzy Networks and Neural Systems Yogyakarta: Graha Ilmu.
• MAVANDI, P., ABBASZADEH, B., EMAMI BISTGANI, Z., BARKER, A.V. and HASHEMI, M., 2021. Biomass, nutrient concentration and the essential oil composition of lavender (Lavandula angustifolia Mill.) grown with organic fertilizers. Journal of Plant Nutrition, vol. 44, no. 20, pp. 3061-3071. http://doi.org/10.1080/01904167.2021.1936037
  » http://doi.org/10.1080/01904167.2021.1936037
• NI, Z.-J., WANG, X., SHEN, Y., THAKUR, K., HAN, J., ZHANG, J.-G., HU, F. and WEI, Z.-J., 2021. Recent updates on the chemistry, bioactivities, mode of action, and industrial applications of plant essential oils. Trends in Food Science & Technology, vol. 110, pp. 78-89. http://doi.org/10.1016/j.tifs.2021.01.070
  » http://doi.org/10.1016/j.tifs.2021.01.070
• OSTADI, A., JAVANMARD, A., AMANI MACHIANI, M., MORSHEDLOO, M.R., NOURAEIN, M., RASOULI, F. and MAGGI, F., 2020. Effect of different fertilizer sources and harvesting time on the growth characteristics, nutrient uptakes, essential oil productivity and composition of Mentha x piperita L. Industrial Crops and Products, vol. 148, pp. 112290. http://doi.org/10.1016/j.indcrop.2020.112290
  » http://doi.org/10.1016/j.indcrop.2020.112290
• PANDEY, D.K., CHAUDHARY, R., DEY, A., NANDY, S., BANIK, R.M., MALIK, T. and DWIVEDI, P., 2020. Current knowledge of cinnamomum species: a review on the bioactive components, pharmacological properties, analytical and biotechnological studies. In: J. SINGH, V. MESHRAM and M. GUPTA, eds. Bioactive natural products in drug discovery (pp. 127-164). Singapore: Springer. https://doi.org/10.1007/978-981-15-1394-7_3
  » https://doi.org/10.1007/978-981-15-1394-7_3
• PEÇANHA, D.A., FREITAS, M.S.M., VIEIRA, M.E., CUNHA, J.M. and DE JESUS, A.C., 2021. Phosphorus fertilization affects growth, essential oil yield and quality of true lavender in Brazil. Industrial Crops and Products, vol. 170, pp. 113803. http://doi.org/10.1016/j.indcrop.2021.113803
  » http://doi.org/10.1016/j.indcrop.2021.113803
• PUTRA, G.B.B. and PRAMESTI, I.G.A.A., 2019. Identification of key variables in sustainable tourism development based on the understanding of accounting students taking tourism business courses. Journal of Accounting Research, vol. 9, no. 2
• RASHID, M., BARI, B.S., YUSUP, Y., KAMARUDDIN, M.A. and KHAN, N., 2021. A Comprehensive review of crop yield prediction using machine learning approaches with special emphasis on palm oil yield prediction. IEEE Access : Practical Innovations, Open Solutions, vol. 9, pp. 63406-63439. http://doi.org/10.1109/ACCESS.2021.3075159
  » http://doi.org/10.1109/ACCESS.2021.3075159
• SHARMA, A., GUMBER, K., GOHAIN, A., BHATIA, T., SOHAL, H.S., MUTREJA, V. and BHARDWAJ, G., 2023. Importance of essential oils and current trends in use of essential oils (aroma therapy, agrofood, and medicinal usage). In G.A. NAYIK and M.J. ANSARI, eds. Essential oils (pp. 53-83). USA: Elsevier. http://doi.org/10.1016/B978-0-323-91740-7.00002-5
  » http://doi.org/10.1016/B978-0-323-91740-7.00002-5
• SINKOVICS, R.R. and ALFOLDI, E.A., 2012. Progressive focusing and trustworthiness in qualitative research: the enabling role of computer-assisted qualitative data analysis software (CAQDAS). MIR. Management International Review, vol. 52, no. 6, pp. 817-845. http://doi.org/10.1007/s11575-012-0140-5
  » http://doi.org/10.1007/s11575-012-0140-5
• SIVAKUMAR, D. and BAUTISTA-BAÑOS, S., 2014. A review on the use of essential oils for postharvest decay control and maintenance of fruit quality during storage. Crop Protection (Guildford, Surrey), vol. 64, pp. 27-37. http://doi.org/10.1016/j.cropro.2014.05.012
  » http://doi.org/10.1016/j.cropro.2014.05.012
• TASHMAN, L.J., 2019. Data science for business and decision making Boca Raton: CRC Press.
• VARGAS, O.S., DE LEÓN ALDACO, S.E., ALQUICIRA, J.A., VELA-VALDÉS, L.G. and NÚÑEZ, A.R.L., 2024. Adaptive Network-Based Fuzzy Inference System (ANFIS) applied to inverters: a survey. IEEE Transactions on Power Electronics, vol. 39, no. 1, pp. 869-884. http://doi.org/10.1109/TPEL.2023.3327014
  » http://doi.org/10.1109/TPEL.2023.3327014
• VESTER, F., 2007. The MICMAC Method: a tool for the analysis of complex systems USA: Springer.
• YOUSEFZADEH, S., MODARRES SANAVY, S.A.M., GOVAHI, M. and KHATAMIAN OSKOOIE, O.S., 2015. Effect of organic and chemical fertilizer on soil characteristics and essential oil yield in dragonhead. Journal of Plant Nutrition, vol. 38, no. 12, pp. 1862-1876. http://doi.org/10.1080/01904167.2015.1061548
  » http://doi.org/10.1080/01904167.2015.1061548
• YULISTYARINI, T., 2020. Reproductive activity patterns of four species of Cinnamomum in Purwodadi Botanic Garden. IOP Conference Series. Earth and Environmental Science, vol. 456, no. 1, pp. 012055. http://doi.org/10.1088/1755-1315/456/1/012055
  » http://doi.org/10.1088/1755-1315/456/1/012055
• ZHAO, G., LIU, S., LU, H., LOPEZ, C. and ELGUETA, S., 2018. Building theory of agri-food supply chain resilience using total interpretive structural modelling and MICMAC analysis. International Journal of Sustainable Agricultural Management and Informatics, vol. 4, no. 3/4, pp. 235. http://doi.org/10.1504/IJSAMI.2018.099236
  » http://doi.org/10.1504/IJSAMI.2018.099236
• ZHAO, L., WANG, Q., HWANG, B.-G. and CHANG-RICHARDS, A.Y., 2024. Analysis of critical factors influencing sustainable infrastructure vulnerabilities using an ISM-MICMAC approach. Engineering, Construction, and Architectural Management, vol. 31, no. 9, pp. 3622-3652. http://doi.org/10.1108/ECAM-07-2022-0690
  » http://doi.org/10.1108/ECAM-07-2022-0690

Publication Dates

• Publication in this collection
  14 Feb 2025
• Date of issue
  2024

History

• Received
  16 Sept 2024
• Accepted
  19 Nov 2024