You are on page 1of 7

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/338169883

Machine Learning Approach for an Automatic Irrigation System in Southern


Jordan Valley

Preprint · October 2020

CITATIONS READS

0 353

3 authors, including:

Anas H. Blasi Rufaydah Al-Huweimel


Mu’tah University Mu’tah University
21 PUBLICATIONS   66 CITATIONS    2 PUBLICATIONS   1 CITATION   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Object-Oriented Software Documentation View project

Cost-Benefit Analysis of Increased Reach Truck Performance in a Materials Handling Warehouse View project

All content following this page was uploaded by Anas H. Blasi on 16 March 2021.

The user has requested enhancement of the downloaded file.


Engineering, Technology & Applied Science Research Vol. X, No. X, 20XX, pp 1

Machine Learning Approach for Automatic


Irrigation System in Southern Jordan Valley
Anas H. Blasi1 Mohammad A. Abbadi2 Rufaydah Al-Huweimel3
Computer Information Systems Computer Science Department Computer Science Department
Department Mutah University Mutah University
Mutah University Al Karak, Jordan Al Karak, Jordan
Al Karak, Jordan

Abstract—The provision of water in the agriculture sector is depends on rains, but for irrigated agriculture, it covers all the
important and has a high priority because the agriculture sector needs of productive trees, vegetables and field crops also.
is the most water consuming sector, and due to the critical irrigated agriculture relies on water sources (from dams,
situation of available water resources in the Hashemite Kingdom groundwater, etc.). The total water resources in the Jordan [4]
of Jordan, attention should be paid to the issue of water needs of shows that the quantities of water in Jordan are somewhat low,
plants and scheduling the appropriate irrigation in order to especially as agricultural land is scattered in the highest
spread the right ways in the management of modern irrigation temperatures’ places (Jordan valleys), which increases the
and effectively transfer to the farmers. This paper objectives are demand due to evaporation and weather factors. Water is
improving the irrigation process to preserve the quality of plants required for the basic growth of the plants, when there is no
and prevent plants drying, and provide irrigation water to the enough water, the quality of the plants will decrease or the
highest extent possible through using the artificial intelligence to
plants will die, so we should look for an irrigation method that
construct a smart irrigation system which control the irrigation
mechanism using the necessary programming for sensing the soil
suits our water supply and plant’s needs. Traditionally, the
moisture, temperature, giving alerts of any change in the first thing that comes to mind is drip irrigation. Drip irrigation
parameters entered as the baseline values for comparison, and is a water-saving system that provides water by slowly
installing system sensors to measure the moisture content in the spraying water into the soil or plant root which contributes to
soil, buried at a depth of 3-5 inches below the areas of roots reducing the spread of herbs and reduces plant diseases and is
spread in green lands, The sensors works on measuring the suitable for all soil types but relies on pipes which needs
humidity and temperature in the soil every ten minutes, it regular maintenance manually to prevent clogging. Another
prevents automatic irrigation process if the humidity is high, and element that plays a major role in agriculture other than water
permitted if the humidity is low. The smart automatic irrigation which is the soil. Soil is a fragile or fragmented surface layer
system model was built using the decision tree (DT) algorithm, that covers the surface of the earth and consists of fragmented
which is one of the machine learning algorithms that trains the rock materials that have been changed because of exposure to
system on part of collected data to build the model that will be environmental, biological and chemical factors [5].
used to examine and predict the other part of the remaining data,
which led to obtaining a system that has an accuracy of 97.86% The soil has several physical properties which mainly
in anticipating the results, which means that it may benefit in affect the growth of the plant, these properties are [5]:
providing a high percentage of irrigation water for the
agricultural sector.
• Soil Strength: represents the aggregation or
arrangement of granules in the soil. The soil strength
Keywords-Irrigation; Agriculture; Artificial Intelligence; Sensor;
is classified according to the following figure to four
Machine Learning; Decision Tree Algorithm. types as shown in Figure 1 below:

I. INTRODUCTION
The Hashemite Kingdom of Jordan follows the climate of
the Mediterranean region. Rainfall from 50 mm per year in the
desert to about 580 mm in the northern highlands snow falls at
short intervals on most mountain highlands in the north,
central and south of the kingdom and are very heavy and
sometimes accumulated [1]. According to this climate
diversity in Jordan, agricultural areas and lands are widely
spread mainly in the Jordan Valley(Northern, Central and
Southern), the total areas of agricultural lands in Jordan in
2017 amounted to 1894324 dunums, 864549.3 dunums from
this area are irrigated agriculture against 102,977,7.7 dunums
rainfed agriculture which depends on rainwater [3], as we
note; the area of rainfed agriculture is larger, that it is because
these areas contains field crops since it is the most plants type
Fig. 1. Soil Strength Classification [5]

www.etasr.com DO NOT ALTER HEADER & FOOTER. THEY WILL BE COMPLETED DURING EDITING
Engineering, Technology & Applied Science Research Vol. X, No. X, 20XX, pp 2

• Organic matter: contribute to the binding of granules in the heat to extreme of cold and from extreme dryness to excessive
soil texture and increase the ability to retain water and good rainfall in addition to irregularity in distribution of rainfall
soil contains 1-6% of the organic matter. Other properties throughout the year, the amount of water in the region and the
such as soil water content, soil color, soil air. quality of the soil , in addition to technical development in all
areas of our life, the idea came to our mind about applying
• Soil Texture: which is the percentage size of mineral grains machine learning to help farmers in plants irrigation without
in the soil from sand, silt and clay and is considered as the time and effort with providing irrigation water quantities and
most important property of the soil. maintain excellent plant production, by designing an irrigation
All these characteristics are closely linked in the system model based on weather and soil type, since the
agricultural process because of its direct effect on the plant, in artificial way of supplying water to the agricultural land at the
the previous Table II, we will describe the soil properties in right moment in an appropriate volume for the proper growth
Gour Al-safi-AlKarak-Jordan [4] as an example of agricultural of the plants in order to get the maximum yields of cultivation
area in this paper. Table I below shows that the soil type in Therefore, the aim of this paper is to create an intelligent
GourAlsafi is a loamy sand soil, which means that the soft, irrigation system that combines the advantages of traditional
coarse and medium grains are close and equal in this type of way advantages (drip irrigation) with a pipe control
soil. This helps make it better than sandy soil with water mechanism for giving the ideal amount of water for the plants,
retention rate and as a result of these homogeneous which contributes to reduce the fungal diseases of the soil,
proportions It helps to be medium water permeability, root rot, and reduce the use of chemical pesticides, which
characterized by moderate ventilation and good fertility rate results in a healthy green land free of diseases. An automatic
and is easier to deal with in terms of ease of cultivation and irrigation system is responsible to water the plant efficiently
productivity of crops. and also providing a system that can reduce the amount of
Back to all the above narratives of the climate quality in manpower needed for watering plant in addition saves
Jordan and weather conditions, which range from extreme of irrigation water.
TABLE I. GOUR ALASAFI SOIL PROPARITIES
Soil Layer Soil Sand Celt % Clay SAT % WP% Field KSAT(m Salinity
Layers Depth (cm) Texture % % Capacity % m/hr.) (MS/cm)

Laye 0-20 Loamy 76.2 20.2 3.6 27.8 2.7 6.5 27.6 9.3
r1 Sand
Laye 20-46 Loamy 84.4 12.9 2.7 29.0 2.1 4.3 42.7 2.7
r2 Sand
Laye 46-87 Loamy 88.9 8.9 2.1 30.0 1.9 2.8 65.6 1.8
r3 Sand
The paper is organized as follows: Section 2 reviews the processing, since if the plant height is short; spread a fertilizer
related work about the use of artificial intelligence on it, if there is no need to fertilization, check the temperature
applications in agriculture and irrigation, section 3 describes and soil moisture, if they less than the set point open the water
the research methodology followed to prepare the data motor on plants. all orders about irrigation controlled by
including data understanding, selecting, transforming, and farmers since the sensor reads the moisture and temperature in
model building. Section 4describes the interpretation and the soil, assesses the plants need for water, and then sends a
evaluation of the results. Finally, section 5 discusses the short message to the farmer to enter the commands and
conclusions and draws the future work. observations in terms of the amount of water and how long it
wants (the possibility of remote control and observation). the
II. RELATED WORK algorithm used is fuzzy logic irrigation algorithm to water the
For several years, there are many papers have been grass and trees based on a set of rules.
published related to the use of artificial intelligence
applications in agriculture and irrigation, through the review According to the paper [7], the idea of the paper was
of many related papers, it has been noticed the absence of designing an automated irrigation system based on scheduling
Arab countries on the research in this field, since there is only irrigation to be twice a day and scheduling the fertilization
one Arab country that has recently started working on this, process to be three times a week. irrigation depends on the soil
which is the United Arab Emirates. The other papers belong to moisture using special sensor to return the level of soil
the United States, India, and Malaysia, so this paper may be moisture and the system will start automatically when the low
considered as one of the first papers that looks at the field of level of moisture, the system will automatically stop when it
artificial intelligence in agriculture in Jordan and Arab reached the suitable level of moisture. In addition to do
countries after the United Arab Emirates. fertigation process by inject fertilizers through pipes of system
three times a week by using a water pump which supplies the
The authors in [6] presents a fully automated drip mixed fertilizer with water to the plants until the desired
irrigation system in India that monitored and controlled using amount of fertilizer is reached. The authors applied the
Raspberry Pi, which is a low cost, credit-card sized computer, Bernoulli’s theorem in order to find the pressure of the water
the idea comes from that the India economy mainly based on supply, the flow rate of water supply and the water losses. The
agriculture and its climate diversity there. This system takes an system has microcontroller as a brain to convert and
image for plant to check its height by making an image

www.etasr.com DO NOT ALTER HEADER & FOOTER. THEY WILL BE COMPLETED DURING EDITING
Engineering, Technology & Applied Science Research Vol. X, No. X, 20XX, pp 3

manipulate the data obtained from soil moisture sensor to get these factors are integrated with GIS for decision-making in
the actual value of humidity. The control unit will instruct the the irrigation process thus obtaining an intelligent irrigation
pump to start working based on the data obtained and the system called fuzzy inference system. A mysterious intelligent
pump will stop working also based on the data obtained by the deduction system based on accurate irrigation knowledge,
microcontroller. In fact, this way served the small home capable of creating guidance maps to control the speed of
gardens. rotation of the irrigation central axis whose inputs are images
processed based on soil characteristics and precise quantities
The other paper [8], concerned about the role of artificial of irrigation using mathematical equations to process
intelligence in improving the agriculture sector and how to fill ambiguous data.
the various agricultures using artificial intelligence and how to
deal with huge amount of data which obtained on a daily time Authors in [10] were discussed the groundwater, its
such as soil reports, plant needs for fertilizer and the use of characteristics and its suitability for irrigating crops through an
robots in improving crop harvesting. The author also focused experiment in the Nanded area /Maharashtra state-India,
on the use of image-based insight generation to conduct a indicating the characteristics of this area in terms of climate,
complete field analysis of agricultural land by combining rainfall and soil quality. An artificial neural network/back
between computer vision technology, the Internet of Things propagation algorithm was used to study the concentration of
(IoT) and drone data to monitor crops and fields, and using all salts and minerals in it, physical and chemical analysis and
of these combinations in disease detection, Field management, design of the optimal model to predict the validity of
Identify the readiness of the crop, Identification of optimal groundwater for irrigation. Fifty representative groundwater
mix for agronomic products, Crop health monitoring, samples were collected from different locations in rural and
Automation techniques in irrigation and enabling farmers by agricultural areas. the used model consists of thirteen input
using AI to train machines based on historical weather neurons (salts and minerals), seven hidden neurons and five
pattern, soil quality and kind of crops to be grown, can output variables to calculate the suitability of groundwater for
automate irrigation then conserve water and benefit farmers in irrigation in the study area. The result is the spatial distribution
managing their water probs. map represented high accuracy of measured and predicted
values. Hence, the proposed ANN model is applicable to
In this paper, India was taken as an example about AI measure the ground water suitability for irrigation in easier
agricultural state, because at present in India, in the state of way. This model can be considered by local development
Andhra Pradesh, Microsoft Corporation is working with authorities for better management of groundwater resources as
farmers rendering farm advisory services using Cortana shown in Figure 3. below:
Intelligence Suite including Machine Learning and Power BI,
it enables in transforming the data into Intelligent Actions. At
the end, the authors talk about challenges in AI adoption in
agriculture like the farming industry is still not having
sufficient service, remains to be underserved, in addition to
extortionate cost of the various cognitive solutions for farming
readily available in the market.
In some agricultural lands, the soil's water needs may vary
from one area to another, although it is the agricultural land
itself, but there are areas where the roots are more common
than others, which affects their moisture. For example, In the
following Figure 2, red color areas need more water than
colored blue, when irrigated these areas should be irrigated Fig. 3. Optimized ANN model to predict the ground water suitability for
according to their variable characteristics, here is called irrigation [10]
Variable rate irrigation (VRI), and if we want to look at the In the [11] paper, the authors (in Iraq- Mosul city /engineering
areas surrounding the circle but outside of it, what is technical college-Mosul garden) present a proposal for a
characteristics? photoelectric irrigation system based on a sensor buried in the
root zone that transmits soil readings to the Arduino board
where the water pump is controlled to be switched ON or OFF
based on sunlight. This system aims to provide water because
irrigation is done automatically according to the level of soil
moisture, as well as save in electrical energy as the pumps are
powered by solar energy through the use of solar cells to
Fig. 2. Spatial variability of irrigation water needs [8] operate them and thus compatible with remote areas where
According to the other paper [9], authors were tackled this there is no electricity grid.
problem by categorizing these areas and everything that [12] This paper aims to build an automatic irrigation system in
applies to “almost” or “about” and "near" all what we consider Indonesia that depends on a programmed chip that controls the
ambiguous in the agriculture sector in general, and the rate of launch of water sprinklers to irrigate crops based on the soil
irrigation in particular, considerate the type of soil, the type of moisture property, whether it is dry or not, these sprinklers
crops, place and time, irrigation method and fuzzy logic. All

www.etasr.com DO NOT ALTER HEADER & FOOTER. THEY WILL BE COMPLETED DURING EDITING
Engineering, Technology & Applied Science Research Vol. X, No. X, 20XX, pp 4

work on electrical energy, this system works using many which is showing the class to which the plant belongs (fruits,
sensors and boards that deal With Arduino technologies and leaves, roots ... etc.), Protected/ exposed which is indicating
screen to display and trace the system outputs, this system is the nature of the environment in which the plant is grown, if it
an extraction from previous research studies in the same field. is in greenhouses or exposed, Land area (dunums) which is the
This project [13] proposes to use photovoltaics to produce area covered by plants, Soil type which describes the nature of
electrical energy that is used as a system source. The humidity the soil to see its ability to retain water and moisture, Month
sensor detects the moisture level in the soil and sends the which is the planting season, Temperature, Morning weather
information to the Arduino controller which in turn sends the humidity which is the degree of humidity in the early morning,
ultrasonic sensor to detect the water level in the tank, The midnight humidity which is the degree of humidity in the
depending on the water level, Arduino controls the water flow late afternoon, The noon weather humidity which is the degree
through the valves, the liquid crystal display (LCD) displays of humidity in the afternoon, Soil moisture which is the degree
information about Soil moisture content and water level, so of moisture in the soil of agricultural lands, Soil salinity which
that information can be conveyed to the end user as short is the percentage of salts in agricultural lands, Evaporation
messages through the GSS module supplied with Arduino, transpiration (m3 / donum / month) which is the amount of
when the humidity level reaches the maximum limit the water that one dunum of agricultural land loses in a month as a
engine stops. Likewise, when the humidity level reaches the result of its evaporation, Average daily water consumption
minimum level the engine starts. (mm) which is the amount of water a plant needs to reach the
Through the previous review of the published papers degree of hydration.
related to the use of artificial intelligence in the field of
agriculture, we note that the potential of machine learning B. Data Selection and Preparation
techniques in this field is multiple and tremendously From the previously collected data, a target data set has
contribute in the agriculture development and improving the created, by selecting a subset of variables (Soil Humidity, Soil
quality of the crops, as it showed that the field is very active Type, Soil Salinity, and Temperature) because it's directly
and that promising new solutions are being developed affects the irrigation process, the selected data was processed,
continuously to meet the challenges facing farmers and
analyzed and correlated using Python programming language.
affecting the economy of countries depends on agriculture.
data preprocessing done for the collected dataset from
Finally, there have been very limited efforts in Arab countries
to use machine learning approaches in agriculture. This work numerical to categorical data, since this model built using the
comes as a beginning. Decision Tree algorithm (DT), “which is a very simple,
understandable, easy to implement and also give a high
III. RESEARCH METHODOLOGY accurate result. It is widely known and used in many
There are many methodologies dealing with data mining businesses to support decision making process and risk
that can be applied to projects related to modern data science. analysis” [15,16]. And this algorithm worked only on
One of the most common of these is the Knowledge Discovery categorical data type.
in Database (KDD) method, which has been adopted as a
methodology for working on this paper. C. Decsion Tree Algorithm
Data mining means extracting useful information from a
large data set which is in its natural state useless for anything
only after making some preprocessing to obtain meaningful
information, thus acquiring knowledge and patterns. In that
sense, Data Mining is also known as Knowledge Discovery
(KD) or Knowledge Extraction (KE). One of the main
purposes of Data Mining is the information gathered from
Fig. 4. Knowledge Database Discovery (KDD) Processes[14]
Data Mining helps to predict hidden patterns, future trends and
As shown in Figuge 4. KDD is the process of discovering behaviors and allowing businesses to take decisions. And it is
useful knowledge from a set of data[14], through a set of also can be applied to any type of data e.g., Data Warehouses,
steps until reaching the results and evaluation as follows: Transactional Databases, Relational Databases, Multimedia
Databases, Spatial Databases, Time-series Databases, World
A. Data Collection Wide Web. In the process of data mining, large data sets are
first sorted, then patterns are identified, and relationships are
The data used in this paper are real data obtained from the
established to perform data analysis and solve problems, this
Directorate of Agriculture of the Southern Jordan Valley/ called Data Classification which is the process of finding a
Department of Agricultural Extension [4], also the National model that describes and distinguishes data classes and
Center for Agricultural Research and Extension/ Soil Survey concepts. Classification is the problem of identifying to which
Project [5], in addition to data about the weather from Taqs of a set of categories (subpopulations), a new observation
Al-Arab website and the Department of Meteorology [1]. Data belongs to, on the basis of a training set of data containing
were compiled as a table in a CSV file contains 1498 records observations and whose categories membership is known, it is
and 15 attributes such as: Plant name which is showing the a process of splitting data set into two groups: first is the
common name of the plants in agricultural areas, Type of plant training dataset: 67% from the data used to train the machine

www.etasr.com DO NOT ALTER HEADER & FOOTER. THEY WILL BE COMPLETED DURING EDITING
Engineering, Technology & Applied Science Research Vol. X, No. X, 20XX, pp 5

classifier for result prediction, then create a general positive (True Positive TP) equals (1225) versus only (26)
classification model using Decision Tree Algorithm to build a predictions that predicted as negative and are actually positive
classifier by making the model learn using the training set (False Positive FP), and the number of predictions which is
available. The model has to be trained for the prediction of predicted as negative and are actually negative (True Negative
accurate results. Second is testing dataset: 33% from the data TN) equals (240) versus only (26) predictions that predicted as
which entered on the model for testing based on training positive and they are actually negative (False Negative FN),
result. here, the constructed model used to predict class labels this shows that how this system is accurate in being able to
and hence estimate the accuracy of the classification rules. predict the needs of plants irrigation based on the readings
provided by the sensors in proportion up to 98%. The
After dividing the data into the previous groups, the
following Table III. also shows the classification and
training process begins by measuring the temperature and evaluation report for irrigation system:
degree of soil moisture through burial sensors buried at a
depth of 3-5 inches below the areas of roots spread in green TABLE III. STRATIFIED CROSS-VALIDATION
lands, in addition to soil salinity feature, giving alerts of any
change in the parameters entered as the baseline values for Accuracy Precision Recall F1 Score MSE
comparison, The sensors works on measuring the humidity
0.0 0.98 0.9 0.98 0.94
and temperature in the soil every 10 minutes, it prevents 0.0213
automatic irrigation process if the humidity is high, and 1.0 0.98 1.0 0.98 0.99
permitted if the humidity is low. in testing phase, the testing The above classification report displays the accuracy,
data set entered as inputs for the designed model then compare precision, recall, F1 scores, MSE, AVG for the model. In
the results between the assumption and the result predicted by order to support easier interpretation and problem detection,
the model. Python programming language is considered the the metrics are defined in terms of TP, FP, TN and FN
most suitable and has been used in this paper because it is predictions. Using this terminology, the metrics are defined as
designed to be extendable with compiler code for proficiency, follows [17,18,19]:
also many tools are available to facilitate Python integration
and software code. However, the prediction model will be • Precision is the ability of a classifier not to label an instance
built using sklearn with some important libraries and methods positive that is actually negative. For each class it is defined
such as accuracy_score, train_test_split, as the ratio of TP to the sum of TP and FP. Said another
DecisionTreeClassifier and panads. way: “for all instances classified positive, what percent was
correct?”, the irrigation system model get a high precision
IV. RESULTS EVALUATION
equal to 0.90.
It is useful to control water consumption, especially in
• Recall is the ability of a classifier to find all positive
agriculture sector, since it is the most sector using the water
instances. For each class it is defined as the ratio of TP to
resources on the earth, also this may affect the quality of
the sum of TP and FN. This model gets 0.98 recall.
production of fruits or vegetables as controlling the amount of
water reduces fungi and plant diseases and thus increases • The F1 score is a weighted harmonic mean of precision and
profits. In this paper, an automatic irrigation system has been recall such that the best score is 1.0 and the worst is 0.0.
designed for agriculture working by measuring soil moisture Generally speaking, F1 scores are lower than accuracy
and temperatures, in addition to some characteristics affecting measures as they embed precision and recall into their
the irrigation process such as soil quality and salinity, using computation. As a rule of thumb, the weighted average of
this system the farmer can follow the crops and monitor them F1 should be used to compare classifier models, not global
without the need for manpower, with saving the time and the accuracy.
cost required for the irrigation process is that the smart
irrigation system is economically feasible and somewhat • Mean square error measures the average squared difference
inexpensive, in addition to the high accuracy that the system between the estimated values and true value. It is a risk
provides. A confusion matrix is often used to describe the function, corresponding to the expected value of the
performance of a classification model on a set of test data for squared error loss, always nonnegative, values close to zero
which the true values are known. The confusion matrix itself are better like the MSE for irrigation system model which is
is relatively simple to understand. The following Table II equal to 0.0213.
shows the system confusion Matrix which is an evaluation • Accuracy: the most commonly used metric to judge a
method for the decision tree model results as follows: model which is the sum of TP with TN divided on the sum
of all classes (TP, TN, FP, FN), this model gets a high
TABLE II. CONFUSION MATRIX
accuracy which equals to 0.98.
N=1497 Predicted “NO” Predicted “YES”
V. CONCLUSION AND FUTURE WORK
Actual “NO” 240 6
The provision of water in the agriculture sector is
Actual “YES” 26 1225
important and has a high priority because the agriculture
sector is the most water consuming sector, and due to the
From the above Table II, we note that the number of critical situation of available water resources in the Hashemite
predictions which is predicted as positive and are actually

www.etasr.com DO NOT ALTER HEADER & FOOTER. THEY WILL BE COMPLETED DURING EDITING
Engineering, Technology & Applied Science Research Vol. X, No. X, 20XX, pp 6

Kingdom of Jordan, attention should be paid to the issue of [9] Willians Ribeiro Mendes, FábioMeneghetti U. Araújo and Salah Er-
water needs of plants and scheduling the appropriate irrigation Raki,Ch3: "Integrating Remote Sensing Data into Fuzzy Control System
in order to spread the right ways in the management of modern for Variable Rate Irrigation Estimates",from Book: "Irrigation; Water
irrigation and effectively transfer to the farmers. Productivity and Operation, Sustainability and Climate Change ",
In this paper, covered the most prominent areas where p.p:272 ,December 18th 2019.
agricultural lands are spread in the Hashemite Kingdom of [10] Vasant Madhav Wagh, Dipak BaburaoPanaskar, Aniket Avinash Muley,
Jordan are the southern Jordan Valley and collected the Shrikant VitthalMukate, Yogesh PopatraoLolage, Manesh Laxman
needed data to design a model for the irrigation process based Aamalawar, "Prediction of groundwater suitability for irrigation using
on soil properties and temperatures, and this system helps in artificial neural network model”. Springer, vol.2, No.196, 10 November
providing the quantities of water needed for irrigation and 2016.
improving productivity, but there are other agricultural regions [11] Safwan A. Hamoodi, Ali N. Hamoodi, Ghanim M. Haydar," Automated
in Jordan that have different characteristics from the southern irrigation system based on soil moisture using arduino board”,Bulletin of
Jordan Valley, especially in temperature, so we may not be Electrical Engineering and Informatics Vol. 9, pp. 870~876, June 2020.
able to generalize the results of this model to all regions, so it [12] Ipin Prasojo, Andino Maseleno, Omar tanane, Nishith Shahu," Design of
is good to improve this model in the next stages to apply this Automatic Watering System Based on Arduino ", Journal of Robotics
model for all agricultural regions in Jordan according to its and Control (JRC) Vol. 1, No. 2, 2020, pp. 55-58 ISSN: 2715-5072,
specifications, develop this model to include scheduling the DOI: 10.18196/jrc.1212, March 2020.
fertilization process for agricultural lands, develop this model [13] Sundar Ganesh C S, Prem Kumar K, Sanjay S, Santhosh H, Gokulnath
to heat the pipes to solve the problem of frost and freezing in M K , "Solar Powered Microntroller Based Irrigation System" , UGC
the winter to raise temperatures, which helps in keeping the Care Journal, ISSN: 2394-3114, Vol-40-Issue-60-March-2020.
plant from withering and shrinking, and using irrigation pipes [14] Techopediawebsite:https://www.techopedia.com/definition/25827/knowl
to improve soil aeration and inject them with organic materials edge-discovery-in-databases-kdd
needed to improve agricultural production combining the
[15] F. LI, Y. LI, and C. WANG, “Uncertain data decision tree classification
automatic irrigation system with image processing through a
algorithm,” Journal of Computer Applications, vol. 29, no. 11, pp. 3092–
set of cameras to monitor plant growth in terms of size and
3095, Dec. 2009.
disease detection. In future work, a different data set will be
[16] Informatic-ar website: https://informatic-ar.com/id3_algorithm/
applied from different regions of The Hashemite Kingdom of
[17] Scikit-learn.orgwebsite: https://scikit-learn.org/
Jordan. In addition to use other machine learning algorithms
[18] Muhammad hasnain, Muhammad fermi pasha, Imran ghani, Muhammad
such as Artificial Neural Networks (ANNs) [20, 21], fuzzy imran, mohammed y. alzahrani, and Rahmatbudiarto. “Evaluating Trust
logic [22], k-NN algorithm [23], and other data mining Prediction and Confusion Matrix Measures for Web Services Ranking”.
techniques [24, 25] then compare the results with DT results. IEEE. Access. May 2020.
[19] Mohammad Hossin, Sulaiman M.N.”A Review on Evaluation Metrics
REFERENCE
for Data Classification Evaluations”. International Journal of Data
[1] Taqs Al-Arab and the Department of Meteorology website:
Mining & Knowledge Management Process 5(2):01-11. March 2015.
https://www.arabiaweather.com/
[20] Anas H. Blasi, “Performance increment of high school students using
[2] Dr. Anas Blasi Al Refae / Faculty of Information Technology-Mutah
ANN model and SA algorithm”. Journal of Theoretical and Applied
University/ Advanced Artificial Intelligence course slides. Geeks for
Information Technology 95(11):2417-2425. 2017.
Geeks: https://www.geeksforgeeks.org/
[21] Rawabi A Aroud, Anas H. Blasi, Mohammed Alsuwaiket. “Intelligent
[3] Ministry of Agriculture-Directorate of Agriculture of the Southern
Risk Alarm for Asthma Patients using Artificial Neural Networks”.
Jordan Valley - Department of Agricultural Extension - Gour Al-
International Journal of Advanced Computer Science and
Safi/Al-Karak
Applications,Vol. 11 No. 3, 95-100, 2020
[4] The National Center for Agricultural Research and Extension / Soil [22] Anas H. Blasi, “Scheduling food industry system using fuzzy logic”.
Survey Project -Gour Al-Safi/Al-Karak Journal ofTheoretical and Applied InformationTechnology 96(19):6463-
[5] Raymond R Weil ,Nyle C Brady ,"The Nature and Properties of Soils. 6473. 2018.
15th edition", 2002 [23] MohammadAl Lababede, Anas H. Blasi, Mohammed A. Alsuwaiket.
[6] Suprabha Jadhav, Shailesh Hambarde,"Android based Automated “Mosques Smart Domes System using Machine Learning Algorithms”.
Irrigation System using Raspberry Pi", International Journal of Science International Journal of Advanced Computer Science and Applications,
Vol. 11, No. 3, 2020.
and Research (IJSR): Vol 5 Issue 6 pID: NOV164836,India, June 2016.
[24] Mohammed A. Alsuwaiket, Anas H. Blasi, khawlaAltarawneh.
[7] A.I.Norazizan,M.A.E.Muzni, M.L.H.Ramsaid, M.Z.Ismail,
“Refining Student Marks based on Enrolled Modules’ Assessment
M.A.D.Mazalan,"DESIGN AND DEVELOPMENT OF THE Methods using Data Mining Techniques”. Engineering, Technology and
IRRIGATION SYSTEM FOR PLANTING ( PART 1 ) " , Technical Applied Science Research. Vol. 10(No. 1):5205-5010. 2020.
Report- Research Gate, DOI: 10.13140/RG.2.1.4088.1529, Malaysia,30 [25] Mohammed A. Alsuwaiket, Anas H. Blasi, Ra'Fat Al-Msie'deen.
May 2016. “Formulating Module Assessment for Improved Academic Performance
[8] V. Dharmaraj and C. Vijayanand,"Artificial Intelligence (AI) in Predictability in Higher Education”. Engineering, Technology and
Applied Science Research. 9(3):4287-4291. 2020.
Agriculture", International Journal of Current Microbiology and Applied
Sciences,Vol. 7, No. 12, pp. 2122-2128 , December 2018.

www.etasr.com DO NOT ALTER HEADER & FOOTER. THEY WILL BE COMPLETED DURING EDITING

View publication stats

You might also like