Institute of Science and Technology

TRIBHUVAN UNIVERSITY
AMRIT SCIENCE CAMPUS
Lainchaur, Kathmandu

Project Report on
Crop Recommendation System
With integrated crop cure and marketplace functionality

Prepared Under the Supervision of

Rakesh Kumar Bachchan
Lecturer
Department of Computer Science and Information Technology
Amrit Science Campus

Submitted By:
Ashish Khatri 23123/076
Kritan Sitaula 23149/076
Nischal Kafle 23165/076

Submitted To:
Tribhuvan University
Institute of Science and Technology

Date: __________________
 AMRIT CAMPUS
Affiliated to Tribhuvan University

SUPERVISOR’S RECOMMENDATION

I hereby recommend that this project prepared under my supervision by Ashish Khatri
(23123/076), Kritan Sitaula (23149/076), and Nischal Kafle(23165/076) entitled “Crop
Recommendation System – with integrated cure recommendation and marketplace” in partial
fulfillment of requirement of the degree of BSc. In Computer Science and Information
Technology (B.Sc. CSIT) be processed for evaluation.

_________________________________
Mr. Rakesh Kumar Bachan
Supervisor
 AMRIT CAMPUS
Affiliated to Tribhuvan University

CERTIFICATE OF APPROVAL

This is to certify that the project report entitles “Crop Recommendation System – with
integrated Cure recommendation and marketplace” submitted by:
Ashish Khatri 23123/076
Kritan Sitaula 23149/076
Nischal Kafle 23165/076
are bonafide students of this institute and the work has been carried out by them under the
supervision of Mr. Rakesh Kumar Bachhan and it is approved for the partial fulfillment of the
requirement of Tribhuvan University, for the award of the degree of B.Sc. Computer Science
and Information technology.

HOD, Deparment of CSIT Supervisor

Amrit Campus Mr. Rakesh Kumar Bachan

Internal Examiner External Examiner

Amrit Campus IOST, Tribhuvan University
 ACKNOWLEDGEMENT
We would like to express our sincere gratitude to all those who have contributed to the completion of
this project report.
First and foremost, we extend our heartfelt thanks to Mr. Rakesh Kumar Bachhan, our project
supervisor, for his invaluable guidance, support, and mentorship throughout the duration of this project.
His expertise, encouragement, and constructive feedback have been instrumental in shaping the
direction and success of our work.
We are deeply thankful to our faculty members and academic advisors for their unwavering support and
encouragement throughout the project. Their insights, expertise, and encouragement have been
invaluable in guiding us through the various stages of the project.
We are grateful to our friends and family members for their unwavering support, encouragement, and
understanding during the challenging periods of this project. Their words of encouragement and belief
in our abilities have been a source of strength and motivation.
Last but not least, we express our gratitude to all the individuals and organizations who have directly
or indirectly contributed to the success of this project. Your support and collaboration have been
invaluable, and we are deeply appreciative of your contributions.
Thank you all for being part of this journey and for your invaluable support and encouragement.
 ABSTRACT
Our project presents the development of a comprehensive software solution aimed at optimizing various
stages of the farming cycle. Our software encompasses functionalities like soil analysis, crop selection,
procurement, plant care recommendations, and direct-to-consumer marketing.
The first stage of the farming cycle involves users providing input on soil nutrients (Nitrogen,
Phosphorous, Potassium, pH) and climatic conditions (temperature, humidity, rainfall). Using a
Random Forest classification algorithm, the software predicts suitable crops for cultivation under the
given conditions, allowing farmers to make informed decisions regarding crop selection. Subsequently,
farmers can directly order seeds online through the software platform.
In the next stage, the software assists farmers in managing plant health by providing recommendations
based on soil nutrient levels and crop type. This includes guidance on fertilization and pest management
strategies tailored to the specific needs of the crops.
Finally, the software facilitates direct-to-consumer marketing by enabling farmers to sell their harvested
produce online. Through an integrated e-commerce platform and powerful Content Management
System (CMS), farmers can manage their online store and facilitate transactions with consumers.
Overall, the software solution aims to empower farmers with data-driven decision-making tools,
streamline agricultural processes, and enhance market connectivity, ultimately contributing to increased
productivity, sustainability, and profitability within the agricultural sector.
 Table of Contents

SUPERVISOR’S RECOMMENDATION ............................................................................................... i

CERTIFICATE OF APPROVAL ............................................................................................................ ii
ACKNOWLEDGEMENT ..................................................................................................................... iii
ABSTRACT ........................................................................................................................................... iv
Table of Contents .................................................................................................................................... v
List of Tables......................................................................................................................................... vii
List of Figures ...................................................................................................................................... viii
Chapter 1: Introduction ........................................................................................................................... 1
1.1 Introduction ................................................................................................................................... 1
1.2 Problem Statement ........................................................................................................................ 1
1.3 Objectives ..................................................................................................................................... 2
1.4 Scope and Limitation .................................................................................................................... 2
1.5 Development Methodology .......................................................................................................... 2
Chapter 2: Background Study and Literature Review ............................................................................ 3
2.1 Background Study ......................................................................................................................... 3
2.1.1 Supervised and Unsupervised Learning ................................................................................. 3
2.2 Literature Review.......................................................................................................................... 3
2.3 Study of Existing Work ................................................................................................................. 4
Chapter 3: System Analysis .................................................................................................................... 5
3.1 Requirement Analysis ................................................................................................................... 5
3.2 Feasibility Analysis ....................................................................................................................... 6
3.3 Analysis ......................................................................................................................................... 8
Chapter 4: System Design ..................................................................................................................... 12
4.1 Design ......................................................................................................................................... 12
4.2 Algorithm Detail ......................................................................................................................... 16
Random Forest Classifier .............................................................................................................. 16
CART Algorithm ........................................................................................................................... 16
Chapter 5: Implementation and Testing ................................................................................................ 18
5.1 Implementation ........................................................................................................................... 18
5.1.1 Tools Used............................................................................................................................ 18
5.1.2 Implementation Details of Modules ..................................................................................... 19
5.2 Testing ......................................................................................................................................... 24
5.2.1 Test Cases for Unit Testing .................................................................................................. 24
5.2.2 Test Cases for System Testing .............................................................................................. 27
5.3 Result Analysis............................................................................................................................ 30
Chapter 6: Conclusion and Future Recommendation ........................................................................... 31
6.1 Conclusion .................................................................................................................................. 31
6.2 Future Recommendation ............................................................................................................. 31
References ................................................................................................................................................ i
Appendices............................................................................................................................................... i
 List of Tables
Table 1: Unit Testing of Authentication and Authorization .................................................................. 24
Table 2: Unit test of Banner Management ............................................................................................ 25
Table 3: Unit Testing of Product module .............................................................................................. 25
Table 4: Unit Testing of Seed module ................................................................................................... 26
Table 5: Unit testing of User Module.................................................................................................... 26
Table 6: Unit testing of cart module...................................................................................................... 26
Table 7: Unit testing of order module ................................................................................................... 27
Table 8: Unit testing of transaction module .......................................................................................... 27
 List of Figures
Figure 1: Use Case Diagram .................................................................................................................. 5
Figure 2: Gantt Chart .............................................................................................................................. 7
Figure 3: Activity Diagram ..................................................................................................................... 8
Figure 4: ER Diagram ............................................................................................................................. 9
Figure 5: DFD Level 0 .......................................................................................................................... 10
Figure 6: DFD level 1 for admin........................................................................................................... 10
Figure 7: DFD1 for farmer .................................................................................................................... 11
Figure 8: DFD 1 for customer ............................................................................................................... 11
Figure 9: Database Design .................................................................................................................... 12
Figure 10: Login form ........................................................................................................................... 13
Figure 11: Registration Form ................................................................................................................ 13
Figure 12: Crop Recommendation form ............................................................................................... 13
Figure 13: Create form .......................................................................................................................... 14
Figure 14: Checkout form ..................................................................................................................... 14
Figure 15: Dialog box ........................................................................................................................... 14
Figure 16: Landing Page Interface ........................................................................................................ 15
Figure 17: Content Management System Interface............................................................................... 15
Figure 18: Random Forest Classification.............................................................................................. 16
Figure 19: Sign Up page .......................................................................................................................... i
Figure 20: Login Page .............................................................................................................................. i
Figure 21: Landing Page ......................................................................................................................... ii
Figure 22: Crop Recommendation Form ................................................................................................ ii
Figure 23: Crop Cure recommendation form ......................................................................................... iii
Figure 24: Marketplace .......................................................................................................................... iii
Figure 25: Farmer CMS dashboard........................................................................................................ iv
Figure 26: Admin CMS dashboard ........................................................................................................ iv
 Chapter 1: Introduction
1.1 Introduction
For modernizing agriculture, we need to make the fusion of technology and traditional farming
practices. We are introducing the software tailored to address the diverse needs of farmers by data-
driven decision-making and integration of agricultural processes. Our college project aims to
bridge the gap between traditional farming methodologies and technology through the
development of an innovative software solution.
Our software is engineered to address three key stages of the farming cycle: planning, cultivation,
and marketing.

With integrated ecommerce With Crop Recommendation

functionality System
Marketing Planning

Cultivation

With Cure Recommendation

System

1. The first stage of our software allows users to input data regarding soil nutrients, such as
Nitrogen, Phosphorous, Potassium, and pH levels, alongside climatic parameters including
temperature, humidity, and rainfall patterns. Utilizing a Random Forest classification
algorithm, the software intelligently analyzes this data to recommend suitable crops for
cultivation, tailored to the specific conditions of the farm using huge dataset. Furthermore, our
software also simplifies online ordering of seeds for the recommended crops.
2. Our software extends support to farmers during the cultivation phase by providing plant care
recommendations. By inputting soil nutrient levels and the name of the cultivated crop,
farmers can receive tailored suggestions for optimal fertilization and pest management
strategies, optimizing crop health and productivity.
3. Moreover, upon harvesting, our software facilitates direct online sales of farm produce to
consumers through a user-friendly e-commerce platform. By cutting out intermediaries and
enabling direct-to-consumer transactions, farmers can establish stronger connections with
consumers, enhance market reach, and command better prices for their produce.

1.2 Problem Statement

Current agricultural practices suffer various problems. Among which, lack of expertise in crop
selection, lack of expertise during cultivation and lack of market to sell produce are some. We
have tried to address these issues in our software with three different software solution integrated
in one package. A unified software solution is needed to integrate soil analysis, crop selection,
procurement, plant care recommendations, and direct-to-consumer marketing, addressing these
challenges and empowering farmers with streamlined processes and enhanced market
connectivity.
1.3 Objectives
The objectives of the project can be stated as:
1. To develop the software solution to address planning, cultivation and marketing of farm
produce.
2. To promote data driven decision in agricultural practices.
3. To motivate farmers in ecommerce through powerful ecommerce tools and CMS.
4. To blend traditional farming practices and modern technological tools.

1.4 Scope and Limitation

1.4.1 Scope
As a web-based system, the system will be easily accessible to those with a device supporting
a browser and an Internet Connection. The software will allow users to input soil nutrient
data (Nitrogen, Phosphorous, Potassium, pH) and climatic conditions (temperature,
humidity, rainfall), which will be processed using advanced algorithm. The software will
facilitate direct online ordering of recommended crop seeds from authorized suppliers,
ensuring timely access to quality inputs for farmers. Upon cultivation, the software will
provide personalized plant care recommendations based on soil nutrient levels (N, P, K) and
crop name, helping farmers in optimizing crop health and productivity. Farmers will be
empowered to sell their harvested produce directly to consumers through an online platform
integrated with a robust Content Management System (CMS), facilitating seamless
transactions and market connectivity.
1.4.2 Limitation
Some of the limitation of this software solution is listed below:
1. The accuracy of crop recommendation and cure recommendation highly relies on
the input data regarding NPK, pH, temperature, humidity and rainfall. Variability
and inadequacy in data quality highly impacts the accuracy.
2. Cure recommendation generated by this system are based on generalized guidelines
and may not address unique pest or disease specific to crop or region.
3. System doesn’t take account for specific regional or microclimatic variations that
could influence crop suitability and management practices.
4. Raising technological literacy among farmers, making digital infrastructure
accessible to farmers and comprehensive support and training to them are
challenging.

1.5 Development Methodology

For this project we adopted XP Methodology (Extreme Programming). With a small team of three,
pair programming can be particularly effective. We worked together in pairs, alternating roles
between writing code and reviewing, thereby promoting knowledge sharing, collaboration, and
higher code quality.
We practiced Continuous Integration (CI) by frequently integrate our work in shared git repository
(link: https://github.com/CSIT-final-year-project/Godmel). We practiced TDD (Test Driven
Development) and ensure software meets specified requirements and remains maintainable over
the time of the project.
 Chapter 2: Background Study and Literature Review
2.1 Background Study
Machine learning is an application of artificial intelligence (AI) that gives systems the ability to
automatically learn and evolve from experience without being specially programmed by the
programmer. The process of learning begins with observations or data, such as examples, direct
experience, or instruction, in order to look for patterns in data and make better decisions in the
future based on the examples that we provide. The main aim of machine learning is to allow
computers to learn automatically and adjust their actions to improve the accuracy and usefulness
of the program, without any human intervention or assistance.

2.1.1 Supervised and Unsupervised Learning

Machine learning techniques can be broadly categorized into the following types:
Supervised learning takes a set of feature/label pairs, called the training set. From this
training set the system creates a generalized model of the relationship between the set of
descriptive features and the target features in the form of a program that contains a set of
rules. The objective is to use the output program produced to predict the label for a previously
unseen, unlabelled input set of features, i.e. to predict the outcome for some new data. Data
with known labels, which have not been included in the training set, are classified by the
generated model and the results are compared to the known labels. This dataset is called the
test set. The accuracy of the predictive model can then be calculated as the proportion of the
correct predictions the model labeled out of the total number of instances in the test set.
Unsupervised learning takes a dataset of descriptive features without labels as a training
set. In unsupervised learning, the algorithms are left to themselves to discover interesting
structures in the data. The goal now is to create a model that finds some hidden structure in
the dataset, such as natural clusters or associations. Unsupervised learning studies how
systems can infer a function to describe a hidden structure from unlabeled data. The system
does not figure out the right output, but it explores the data and can draw inferences from
datasets to describe hidden structures from unlabeled data. Unsupervised learning can be
used for clustering, which is used to discover any inherent grouping that are already present
in the data. It can also be used for association problems, by creating rules based on the data
and finding relationships or associations between them.
For our system we are using supervised learning algorithm.

2.2 Literature Review

[1] In paper “Smart Management of Crop Cultivation using IoT and Machine Learning” by
Archana Gupta, Dharmil Nagda, Pratiksha Nikhare, Atharva Sandbhor, author has proposed a
smart system that can assist farmers in crop management by considering sensed parameters
(temperature, humidity) and other parameters (soil type, location of farm, rainfall) that predicts
the most suitable crop to grow in that environment.
[2] In paper “Crop Recommendation on Analyzing Soil Using Machine Learning” by Anguraj.Ka,
Thiyaneswaran.Bb, Megashree.Gc, Preetha Shri.J.Gd, Navya.Se, Jayanthi. Jf, author has proposed
new technologies include the use of Internet of Things (IOT) and Machine Learning. The real time
data from the field area can be collected using IOT system. The collected data from the field area
is fed to the trained model. The trained model then makes the predictions using the data. The result
produced by the model greatly helps is sowing the suitable crops in the particular field area.
[3] In Reference paper “Classification of Soil and Crop Suggestion using Machine Learning
Techniques”, by A. Mythili, determines a model is proposed for predicting the soil type and
 suggest a suitable crop that can be cultivated in that soil. The model has been tested using various
machine learning algorithms such as KNN, SVM and logistic regression. The accuracy of the
present model is maximum than the existing models.
[4] In “IOT based Crop Recommendation, Crop Disease Prediction and Its Solution” Rani
Holambe, Pooja Patil, Padmaja Pawar, Saurabh Salunkhe, Mr. Hrushikesh Joshi, proposed a
system would assist the farmers in making an informed decision about which crop to grow
depending on a variety of environmental and geographical factors. The ML and IoT based
suggestions will significantly educate the farmer and help them minimize costs and make strategic
decisions by replacing intuition and passed-down knowledge with far more reliable data-driven
ML models. This allows for a scalable, reliable solution to an important problem affecting
hundreds of millions of people.

2.3 Study of Existing Work

Different e-commerce platform has been developed in Nepal to sell farmer’s good online such as
Krishi COOP Bazaar. Also, different machine learning model has been developed in for prediction.
Different work such as Regularized Greedy Forest to see an appropriate crop sequence at a given
time stamps. Other research used historical records of meteorological data as training set where
model is trained to spot climate that are deterrent for the assembly of crops which efficiently
predicts the yield of crops on the idea of monthly weather patterns.
Our system implements both e-commerce functionalities and machine learning models for crop
recommendation and soil analysis. We also believed that not only either soil nutrients or weather
data should be taken consideration for training the model but all the appropriate parameters such
as temperature, rainfall, geography and soil condition to predict crop suitability.
 Chapter 3: System Analysis
3.1 Requirement Analysis
Functional Requirements:
Our system allows three kinds of user privileges. One is admin, another farmers and other is
customer. The functionality available for each privileges are represented with Use case diagram
below.

Figure
1: Use Case Diagram

• Admin
1. Use ‘Crop Recommendation System” and “Crop Cure Recommendation System”
2. Create seeds that is recommended for purchase to any user who obtains prediction
of the crop
3. Modify/Delete the created seed
4. Create farm products to list in market place from where any user can make an
order
5. Modify/delete products
6. Create another admin
7. Modify/Delete users
 8. Manage Orders
• Farmer
1. Use ‘Crop Recommendation System” and “Crop Cure Recommendation System”
2. Make order for seeds of crop recommended by crop recommendation system
3. Create farm products to list in market place from where any user can make an
order
4. Modify/delete products
5. Modify/Delete users
6. Manage Orders
• Customer
1. Create Account
2. Make orders for seeds or farm products

Hardware & Software Requirements

For this project, we require laptop with support of HTML5, CSS, and other implementation tools
required for the project. The minimum system requirements for the laptop include: i5 11 th
generation, 16GB RAM, 512 SSD, Windows 11 OS.
The software requirements for the project includes powerful IDE that is Visual Studio Code,
Jupyter Notebook for training models, MongoDB for database, Microsoft edge browser, postman
for API testing and many other open source programs. Draw.io and MS Word are also needed fro
documentation.
Non-Functional Requirements
The non-functional requirements of our software are listed below:
1. Secured
2. Easily accessible
3. Responsive Design
4. User Friendly
5. High Performance

3.2 Feasibility Analysis

A feasibility study is conducted to assess the practicality and viability of implementing the
proposed software solution for the farming cycle. This study evaluates various aspects, including
technical, economic, operational, and scheduling feasibility, to determine whether the project is
feasible and worth pursuing.
1. Technical Feasibility:
This software solution encompasses complex functionalities such as soil nutrient analysis,
crop classification, e-commerce integration, and data management. We have rich set of
historical data regarding crop nature based on soil nutrients data. We are well familiar with
classification algorithms. We are capable of training the model and integrate into webapp
along with developing ecommerce functionality in web app. So, the project is technically
feasible.
2. Economic Feasibility:
Software requirement for development of this project is available free for community
development. However, hardware requirement might be high to setup IOT as we might
need NPK sensor for soil nutrient, moisture sensor for moisture measurement in soil,
Arduino board, etc. But as for now this project is just a prototype so we are going to take
all these parameters through user and weather data through online free APIs like open
weather map. So, for prototyping, this project is economically feasible.
3. Operational Feasibility:
Operational feasibility assessment indicates that the software solution can be effectively
implemented within the existing agricultural ecosystem. User acceptance testing and
stakeholder engagement activities needs to be done before putting it into practice.
4. Scheduling Feasibility
The project expands over a period of about 4 months. The project schedule is feasible.
Schedule plan is shown in the Gantt chart below.

Figure 2: Gantt Chart

3.3 Analysis
Activity Diagram

Figure 3: Activity Diagram

 ER Diagram

Figure 4: ER Diagram

Figure 4: ER Diagram
Data Flow Diagram
1. DFD 0
The context level DFD describes the whole system. Below DFD of the Kishan Saathi App
shows three users can operate the system; Admin, Farmer and Customer.

Figure 5: DFD Level 0

2. DFD 1 for Admin

Figure 6: DFD level 1 for admin

3. DFD 1 for Farmer and customer

Figure 7: DFD1 for farmer

Figure 8: DFD 1 for customer

 Chapter 4: System Design
4.1 Design
Database Design

Figure 9: Database Design

Forms Design
For our system, following forms will be necessary:
1. Login and Registration form
2. Crop recommendation form
3. Cure Recommendation form
4. Checkout form
5. Banner, Seed, Product, User creation form in CMS

The respective design of forms are shown below:

 Figure 10: Login form

Figure 11: Registration Form

Figure 12: Crop Recommendation form

 Figure 13: Create form

Figure 14: Checkout form

Dialog Design

Figure 15: Dialog box

Interface design
There are two different interface design. One for landing pages. Other for Content Management
System.

Figure 16: Landing Page Interface

Figure 17: Content Management System Interface

4.2 Algorithm Detail
Random Forest Classifier
The Random Forest or Random Decision Forest is a supervised Machine learning algorithm used
for classification, regression, and other tasks using decision trees. Random Forests are particularly
well-suited for handling large and complex datasets, dealing with high-dimensional feature spaces,
and providing insights into feature importance. This algorithm’s ability to maintain high predictive
accuracy while minimizing overfitting makes it a popular choice across various domains,
including finance, healthcare, and image analysis, among others.
The Random Forest classifier creates a set of decision trees from a randomly selected subset of
the training set. It is a set of decision trees (DT) from a randomly selected subset of the training
set and then it collects the votes from different decision trees to decide the final prediction.

Figure 18: Random Forest Classification

Each decision tree in the random forest is constructed using a subset of the training data and a
random subset of features introducing diversity among the trees, making the model more robust
and less prone to overfitting.
During the training phase, each tree is built by recursively partitioning the data based on the
features. At each split, the algorithm selects the best feature from the random subset, optimizing
for information gain or Gini impurity. The process continues until a predefined stopping criterion
is met, such as reaching a maximum depth or having a minimum number of samples in each leaf
node.
For creating Decision Tree, we used CART Algorithm. CART stands for Classification and
Regression Tree Algorithm.

CART Algorithm
Classification and Regression Trees (CART) is a decision tree algorithm that is used for both
classification and regression tasks. It is a supervised learning algorithm that learns from labelled
data to predict unseen data.
Tree structure: CART builds a tree-like structure consisting of nodes and branches. The nodes
represent different decision points, and the branches represent the possible outcomes of those
decisions. The leaf nodes in the tree contain a predicted class label or value for the target
variable.
Splitting criteria: CART uses a greedy approach to split the data at each node. It evaluates all
possible splits and selects the one that best reduces the impurity of the resulting subsets. For
classification tasks, CART uses Gini impurity as the splitting criterion. The lower the Gini
impurity, the more pure the subset is. For regression tasks, CART uses residual reduction as the
splitting criterion. The lower the residual reduction, the better the fit of the model to the data.
Pruning: To prevent overfitting of the data, pruning is a technique used to remove the nodes that
contribute little to the model accuracy. Cost complexity pruning and information gain pruning
are two popular pruning techniques. Cost complexity pruning involves calculating the cost of
each node and removing nodes that have a negative cost. Information gain pruning involves
calculating the information gain of each node and removing nodes that have a low information
gain.
The CART algorithm works via the following process:
• The best-split point of each input is obtained.
• Based on the best-split points of each input in Step 1, the new “best” split point is
identified.
• Split the chosen input according to the “best” split point.
• Continue splitting until a stopping rule is satisfied or no further desirable splitting is
available.

CART algorithm uses Gini Impurity to split the dataset into a decision tree. It does that by
searching for the best homogeneity for the sub nodes, with the help of the Gini index criterion.
 Chapter 5: Implementation and Testing
5.1 Implementation
5.1.1 Tools Used
1. Visual Studio Code:
We used VS Code IDE for development of the project. Visual Studio Code (VS Code) is
a free source-code editor developed by Microsoft for Windows, Linux, and macOS. It
offers a wide range of features for developers, including support for debugging, syntax
highlighting, intelligent code completion, code refactoring, and Git integration.
One of the key features of Visual Studio Code is its extensibility, which allows developers
to customize and enhance their coding experience through various extensions available in
the Visual Studio Code Marketplace. These extensions can add support for additional
programming languages, provide new themes, integrate with other tools and services, and
more.

2. Front End Tools:

For designing front end, we used ReactJS. ReactJS is a JavaScript library developed by
Facebook for building user interfaces (UIs) for web applications. It's often used for
creating interactive and dynamic UI components in single-page applications (SPAs) and
mobile applications.
We used React bootstrap to develop frontend components. React Bootstrap is a library
that combines the power of React components with the styling and components of the
Bootstrap framework. Bootstrap is a popular front-end framework for building responsive
and mobile-first websites and web applications. React Bootstrap allows developers to use
Bootstrap's UI components in their React applications, leveraging the advantages of both
technologies.
Additionally, we used other tools for maintaining aesthetics of the software like font-
awesome, styled components, slick-carousel, react-toastify, etc.

3. Back End Tools:

The backend of the project can be divided into two parts. One is machine learning backend
and other is backend of ecommerce functionality.
Backend of machine learning is built on python and integrated into flask application.
Python is widely favored for machine learning due to its simplicity, extensive library
support, vibrant community, flexibility, and seamless integration with other technologies.
Its straightforward syntax makes it accessible to both beginners and experienced
developers, enabling them to focus on solving machine learning problems rather than
grappling with complex language intricacies. Python boasts a rich ecosystem of libraries
tailored for data manipulation and machine learning tasks, such as NumPy, Pandas, and
scikit-learn, facilitating efficient data preprocessing, model building, and evaluation.
Backend of ecommerce functionality was built on NodeJS runtime environment using
Express framework. Node.js is a runtime environment that allows developers to run
JavaScript on the server-side. Express.js is a web application framework for Node.js,
designed for building web applications and APIs.When used together, Node.js with
Express provides a powerful platform for creating server-side applications in JavaScript.
Express simplifies the process of handling HTTP requests and responses, routing,
middleware integration, and more, while Node.js enables high-performance, non-
blocking I/O operations, making it ideal for building scalable and efficient server-side
applications.
 Additionally, JWT was integrated for authorization. Multer was used for file uploading
and handling in server side.

4. Database:
MongoDB database was used as database. MongoDB is a popular open-source NoSQL
database management system that stores data in flexible, JSON-like documents. It's
known for its scalability, flexibility, and performance, making it suitable for a wide range
of applications, from small startups to large enterprises.

5.1.2 Implementation Details of Modules

1. For Crop Recommendation System
Below is a screenshot of first five data of data set:

The function to read the dataset is as:

def read_csv(filename):

data = []

with open(filename, 'r') as file:

header = next(file) # Skip the header row

for line in file:

data.append(line.strip().split(','))

return data

Crop label Dictionary to convert string label into integer:

crop_dict = {
'rice': 1,
'maize': 2,
'jute': 3,
'cotton': 4,
'coconut': 5,
'papaya': 6,
'orange': 7,
'apple': 8,
'muskmelon': 9,
'watermelon': 10,
'grapes': 11,
'mango': 12,
'banana': 13,
'pomegranate': 14,
'lentil': 15,
'blackgram': 16,
'mungbean': 17,
'mothbeans': 18,
'pigeonpeas': 19,
'kidneybeans': 20,
'chickpea': 21,
'coffee': 22
}
This dictionary is used to create a new column for converting string label into numeric value by
the line of code: y=np.array([crop_dict[label] for label in y])

The features and labels are preprocessed by following function:

def preprocess_data(data):

data_array = np.array(data)

X = data_array[:, :-1].astype(float)

y = data_array[:, -1]

# Convert labels to integers using crop_dict

y = np.array([crop_dict[label] for label in y])

return X, y

Now the data is splitted into training and testing data. Train size was was taken 0.8 part of size of
features and test size was 0.2 part.
train_size = int(0.8 * len(features))
X_train, X_test = features[:train_size], features[train_size:]
y_train, y_test = labels[:train_size], labels[train_size:]

Now we create random forest classifier’s object

rf_classifier = RandomForest(n_estimators=100)

The equivalent class for random forest algorithm is

class RandomForest:
def __init__(self, n_estimators=100, max_depth=None, sample_size=None):
self.n_estimators = n_estimators
self.max_depth = max_depth
self.sample_size = sample_size
self.estimators = []

def fit(self, X, y):

for _ in range(self.n_estimators):
tree = DecisionTree(max_depth=self.max_depth)
if self.sample_size:
indices = np.random.choice(len(X), size=self.sample_size, replace=True)
X_sample = X[indices]
y_sample = y[indices]
else:
X_sample = X
y_sample = y
tree.fit(X_sample, y_sample)
self.estimators.append(tree)

def predict(self, X):

predictions = np.array([estimator.predict(X) for estimator in self.estimators])
return np.apply_along_axis(lambda x: np.bincount(x).argmax(), axis=0, arr=predictions)
Then we call the fit function in random forest classifier’s object to train our split train data
rf_classifier.fit(X_train, y_train)

This fit function creates object of Decision tree. The related Decision Tree Class is:
class DecisionTree:
def __init__(self, max_depth=None):
self.max_depth = max_depth

def fit(self, X, y):

self.tree = self._grow_tree(X, y)

def _grow_tree(self, X, y, depth=0):

num_samples, num_features = X.shape
num_classes = len(np.unique(y))

# Stopping criteria
if (self.max_depth is not None and depth >= self.max_depth) or num_classes == 1:
return {'class': np.bincount(y).argmax(), 'num_samples': num_samples, 'depth': depth}

# Find the best split

best_feature = None
best_threshold = None
best_gini = np.inf
for feature in range(num_features):
thresholds = np.unique(X[:, feature])
for threshold in thresholds:
left_indices = np.where(X[:, feature] <= threshold)[0]
right_indices = np.where(X[:, feature] > threshold)[0]
if len(left_indices) == 0 or len(right_indices) == 0:
continue
gini = self._gini_impurity(y[left_indices], y[right_indices])
if gini < best_gini:
best_feature = feature
best_threshold = threshold
best_gini = gini

if best_gini == np.inf:
return {'class': np.bincount(y).argmax(), 'num_samples': num_samples, 'depth': depth}

# Split the dataset

left_indices = np.where(X[:, best_feature] <= best_threshold)[0]
right_indices = np.where(X[:, best_feature] > best_threshold)[0]
left_tree = self._grow_tree(X[left_indices, :], y[left_indices], depth + 1)
right_tree = self._grow_tree(X[right_indices, :], y[right_indices], depth + 1)

return {'feature': best_feature, 'threshold': best_threshold,

'left': left_tree, 'right': right_tree, 'num_samples': num_samples, 'depth': depth}

def _gini_impurity(self, left_y, right_y):

p_left = len(left_y) / (len(left_y) + len(right_y))
p_right = len(right_y) / (len(left_y) + len(right_y))
return p_left * self._gini(left_y) + p_right * self._gini(right_y)

def _gini(self, y):

if len(y) == 0:
return 0
p = np.bincount(y) / len(y)
return 1 - np.sum(p ** 2)
After that fit function of Random Forest call fit function of Decision which will create decision
trees from the training set. After that all the decision trees are appended in the fit function of
Random Forest. Now the model is ready to make predictions. We need new conditions taken as
input from the user to make prediction.
predictions = rf_classifier.predict(condn)

The condn parameter is supply via user input as:

@ app.route('/crop-predict', methods=['POST'])
def crop_prediction():
title = 'Crop Recommendation Module'
print("hello")
if request.method == 'POST':
N = int(request.form['nitrogen'])
P = int(request.form['phosphorous'])
K = int(request.form['potassium'])
ph = float(request.form['ph'])
rainfall = float(request.form['rainfall'])
# temperature = float(request.form['temperature'])
# humidity = float(request.form['humidity'])

city = request.form.get("city")

if weather_fetch(city) is not None:

temperature, humidity = weather_fetch(city)
else:
return json.dumps({"error": "City not found"})

feature_list = [N, P, K, temperature, humidity, ph, rainfall]

single_pred = np.array(feature_list).reshape(1, -1)

prediction = crop_recommendation(single_pred)

response = {
"result": prediction[0],
"message": "Crop Predicted"
}

return json.dumps(response)

2. For Cure Recommendation System

We have a dataset which contains suitable value of N, P, and K for different crops. Below is the
screenshot of few data.
Now, we are taking input of crop name and value of N, P, K from user as:
crop_name = str(request.form['cropname'])

N = int(request.form['nitrogen'])

P = int(request.form['phosphorous'])

K = int(request.form['potassium'])

We extract suitable value of N, P, K as nr, pr, kr from file. Then we find n, p, k which is the
difference between actual N, P, K and suitable nr, pr, kr respectively. Based on that we can
determine which of the parameter is abundance or low in soil.
df = pd.read_csv('Data/fertilizer.csv')

nr = df[df['Crop'] == crop_name]['N'].iloc[0]
pr = df[df['Crop'] == crop_name]['P'].iloc[0]
kr = df[df['Crop'] == crop_name]['K'].iloc[0]

n = nr - N
p = pr - P
k = kr - K
temp = {abs(n): "N", abs(p): "P", abs(k): "K"}
max_value = temp[max(temp.keys())]
if max_value == "N":
if n < 0:
key = 'NHigh'
else:
key = "Nlow"
elif max_value == "P":
if p < 0:
key = 'PHigh'
else:
key = "Plow"
else:
if k < 0:
key = 'KHigh'
else:
key = "Klow"

We create a dictionary of cure for each case of high and low. We markup the dictionaries content
and dumps the final cure recommendation.

response = Markup(str(fertilizer_dic[key]))
res = {
"result": response,
"message": "Recommended cure"
}

return json.dumps(res)
 3. For ecommerce functionality:
No major algorithm was used in this functionality. An express
app was setup in server file and the system was broken down
into modules; auth, banner, cart, dashboard, product, seed, and
user. MVC pattern was followed. Folder were organized into
modular architecture. Each folder of modules contain file for
router, controller, service validator and model. Common
middleware, routes and config are stored in separate folders.
Project variable are controlled from ‘.env’ file.

5.2 Testing
5.2.1 Test Cases for Unit Testing
Table 1: Unit Testing of Authentication and Authorization
Test Description Test Step Expected Result Status
TA1 User Registration Go to sign up page, fill After filling registration pass
test registration form form, user will get
verification link at specified
email
TA2 Verification Check verification email If verified, user will redirect pass
and click on the link to password set page
provided to verify
TA3 Set Password In password set page, Password should be stored pass
create password. Only in database in encrypted
password with Uppercase, form and user should be
Lowercase, Number, redirect into login page.
Symbol and at least 8
characters are allowed.
TA4 Login Provide login credential in For Admin, redirect in pass
login page. ADMIN CMS.
For Farmer, redirect in
FARMER CMS.
For Consumer, redirect in
HOMEPAGE.
TA5 Permission Test Browse Admin Block Admin Dashboard pass
Dashboard URL, while and redirect in homepage
login as User
TA6 Logout Click on logout User should be logged out pass

Table 2: Unit test of Banner Management

Test Description Test Step Expected Result Status
TB1 Banner Create Go to Banner Banner should be created pass
management in sidebar of and displayed in homepage.
admin panel. Click on
create and provide
necessary information to
create banner.
TB2 Banner List Go to Banner There should be list of all pass
Management page the banners created by
logged in user
TB3 Banner Update Click on the yellow pencil The edited information pass
icon button at the last should be reflected.
column of banner listing
and edit the necessary
information
TB4 Banner Delete Click on the red trash icon Confirmation dialog should pass
button at the last column appear and upon confirming
of banner listing the related banner data
should be deleted from
database

Table 3: Unit Testing of Product module

Test Description Test Step Expected Result Status
TP1 Product Create Go to Product Product should be created pass
management in sidebar of and displayed in
admin panel. Click on Marketplace.
create and provide
necessary information to
create product.
TP2 Product List Go to Product There should be list of all pass
Management page the products created by
logged in user
TP3 Product Update Click on the yellow pencil The edited information pass
icon button at the last should be reflected.
column of product listing
and edit the necessary
information
TP4 Product Delete Click on the red trash icon Confirmation dialog should pass
button at the last column appear and upon confirming
of product listing the related product data
should be deleted from
database
 Table 4: Unit Testing of Seed module
Test Description Test Step Expected Result Status
TS1 Seed Create Go to Seed management Seed should be created and pass
in sidebar of admin panel. displayed as product to add
Click on create and on cart when predicted by
provide necessary crop recommendation
information to create system.
product.
TS2 Seed List Go to Seed Management There should be list of all pass
page the seeds created by logged
in user.
TS3 Seed Update Click on the yellow pencil The edited information pass
icon button at the last should be reflected.
column of seed listing and
edit the necessary
information.
TS4 Seed Delete Click on the red trash icon Confirmation dialog should pass
button at the last column appear and upon confirming
of seed listing. the related seed data should
be deleted from database.

Table 5: Unit testing of User Module

Test Description Test Step Expected Result Status
TU1 Create Admin Login as admin and go to New user with role admin pass
user management. Click should be created.
on create admin to create
a new admin.
TU2 User List Go to User Management There should be list of all pass
page the Users.
TU3 User Update Click on the yellow pencil The edited information pass
icon button at the last should be reflected.
column of user listing and
edit the necessary
information.
TU4 User Delete Click on the red trash icon Confirmation dialog should pass
button at the last column appear and upon confirming
of user listing. the related user data should
be deleted from database.

Table 6: Unit testing of cart module

Test Description Test Step Expected Result Status
TAC1 Product Added to In Marketplace, click add A modal for selecting pass
cart to cart button of any quantity will popped and the
product product should be added to
cart afterward.
TAC2 Product checkout In cart, select all the Total should be displayed, pass
product to checkout, enter and after clicking checkout
the delivery address and order should be updated in
click checkout seller’s profile
 Table 7: Unit testing of order module
Test Description Test Step Expected Result Status
TO1 Order Listing Go to order management All the new order should be pass
page listed
TO2 Mark Dispatched At the last column, click The order should be pass
the mark dispatched completed and completed
button order should be displayed in
transaction page

Table 8: Unit testing of transaction module

Test Description Test Step Expected Result Status
TT1 Transaction Listing Go to transaction page All the completed order pass
should be displayed

5.2.2 Test Cases for System Testing

To test the proper functioning of the system, we take a test cases as:
1. We logged in as admin user and created new banner, seeds, product and new admin account
from default admin account. Changes is expected to reflect.
Following is the result. Few Banners, Seeds, Products and new admin is created from admin
CMS.
Banner:.

Seed:

Product:

New Admin:
2. A new farmer user is signed up with the required credentials. Then the user first uses the crop
recommendation system. A random data will be provided to the crop recommendation form.
To get the crop recommendation. User must be able to order the recommended crop from the
recommendation. The order must be shown in admin panel. Then farmer user will test cure
recommendation system. A test data will be fed and the result for curing the crop should be
obtained. From farmer CMS, we will create few products, which must be displayed in Farmer
Marketplace.
Result:
New Farmer account:

Crop recommendation:

Cure recommendation:

Product created in farmer account

 Order list:

3. A new customer user is signed up with the required credentials. Then the customer user will
navigate to farmer market place. Customer user must be able to place order. The order should
be updated in farmer’s panel.

4. Additionally, paypal integration is also tested while ordering.

5.3 Result Analysis
Analysis of crop recommendation system:
From above user input data, system receive following feature list:
"feature_list": [120, 69, 100, -1.03, 53, 6.0, 200.0]
Where correspondence mapping is:
feature_list = [N, P, K, temperature, humidity, ph, rainfall]

The feature_list is fed into the model which has 100 decision tree. Each decision tree derive it’s
own conclusion. The conclusion with highest majority which is “apple” is resulted.

The confidence level of the certainity of the prediction is calculated as

Analysis of cure recommendation system:

Provided feature: [‘grapes’, 50, 60, 70]
In data set, suitable value of n, p, k is as follows:

The module calculate nr, pr, kr as:

nr = 50-20=30
pr = 60-125=-65
kr = 70-200 = -130
By taking absolute value of each, maximum value is kr. So potassium content is needed to be cure
first. Since kr<0, we can assume potassium content is low in the soil. The corresponding dictionary
recommendation for low potassium is as follows:

Thus, we get result as:

 Chapter 6: Conclusion and Future Recommendation
6.1 Conclusion
The development of the proposed software solution for the farming cycle represents a significant
step towards revolutionizing agricultural practices and empowering farmers with data-driven
decision-making tools. By seamlessly integrating soil analysis, crop selection, procurement, plant
care recommendations, and direct-to-consumer marketing functionalities, the software addresses
key challenges faced by farmers and enhances productivity, sustainability, and market access
within the agricultural sector.
Through the implementation of advanced algorithms such as the Random Forest classification
algorithm, the software effectively predicts suitable crops based on soil nutrients and climatic
conditions, enabling farmers to make informed decisions regarding crop selection. The direct
online ordering feature further streamlines the procurement process, providing farmers with
convenient access to quality seeds tailored to their specific requirements.
Additionally, the software's plant care recommendation module offers valuable guidance to
farmers throughout the cultivation process, facilitating optimal fertilization and pest management
strategies based on soil nutrient levels and crop type. The integration of e-commerce capabilities
enables farmers to sell their produce directly to consumers, eliminating intermediaries and
fostering stronger connections between producers and consumers.

6.2 Future Recommendation

In future, the crop recommendation system and cure recommendation system can be automated
with used of sensors. NPK sensor is available in the market which measures Nitrogen,
Phosphorous and Potassium content of the soil. Soil moisture sensor with pH indicator is available
in the market to measure humidity and pH value of the soil. Other data can be fetched from
different API. The wi-fi module can be attach in Arduino board along with other sensor to get real
time data from the field and get real time recommendation without the need of inputting soil data
by user. This will not only made the process easier but also provide real time data.
Other future recommendations are as follows:
a. Enhanced Data Analytics: Continuously improve the software's data analytics capabilities by
incorporating additional factors such as soil texture, crop history, and environmental factors
to further enhance crop prediction accuracy and recommendation precision.
b. Mobile Application Development: Develop a mobile application version of the software to
provide farmers with on-the-go access to essential features and functionalities, enhancing
usability and convenience.
c. Expansion of Market Connectivity: Explore opportunities to expand the software's market
connectivity beyond direct-to-consumer sales, such as partnerships with local markets,
restaurants, and food delivery services to broaden the reach and impact of farmers' produce.
d. Integration with IoT Devices: Incorporate Internet of Things (IoT) devices such as soil
sensors, weather stations, and drones to enable real-time data collection and monitoring,
facilitating proactive decision-making and resource optimization.
e. Community Engagement and Training: Organize workshops, training sessions, and
community outreach programs to educate farmers on the benefits and usage of the software,
fostering adoption and maximizing its impact on agricultural communities.
f. Continuous Improvement and Updates: Implement a feedback mechanism to gather user
feedback and suggestions for software improvements, ensuring ongoing refinement and
enhancement of features based on user needs and evolving industry trends.
 References
[1] 2019 IEEE “Smart Management of Crop Cultivation using IoT and Machine Learning” Archana
Gupta, Dharmil Nagda, Pratiksha Nikhare, Atharva Sandbhor
[2] “Crop Recommendation on Analyzing Soil Using Machine Learning” Anguraj.Ka,
Thiyaneswaran.Bb, Megashree.Gc, Preetha Shri.J.Gd, Navya.Se, Jayanthi. Jf, 2020.
[3] “Classification of Soil and Crop Suggestion using Machine Learning Techniques”, A. Mythili, IEEE
2019.
[4] “IOT based Crop Recommendation, Crop Disease Prediction and Its Solution” Rani Holambe, Pooja
Patil, Padmaja Pawar, Saurabh Salunkhe, Mr. Hrushikesh Joshi, 2019 IRJET
[5] Kaggle: https://www.kaggle.com/datasets
 Appendices

Figure 19: Sign Up page

Figure 20: Login Page

 Figure 21: Landing Page

Figure 22: Crop Recommendation Form

Figure 23: Crop Cure recommendation form

Figure 24: Marketplace

Figure 25: Farmer CMS dashboard

Figure 26: Admin CMS dashboard