VOICE TO VISUAL STORY GENERATOR SYSTEM

PROJECT REPORT

Submitted by the Bharathidasan University in fulfilment of the

requirement for the award of the degree

BACHELOR OF COMPUTER SCIENCE

Submitted By

[Link]

(REG NO:23054021802112017)

Guided by

Mrs E. Sumathi BCA., [Link]., [Link]

Assistant Professor

Department of Computer Science & Computer Applications

KARUR VELALAR COLLEGE OF ARTS AND SCIENCE FOR WOMEN KARUR

Affiliated to Bharathidasan University| An ISO 21001:2018&9001:2015 Certified

Institution

BHARATHIDHASAN UNIVERSITY, TIRUCHIRAPALLI_620024

MARCH-2026
CERTIFICATE
 CERTIFICATE
This is to certify that the dissertation, entitled "VOICE TO VISUAL STORY GENERATOR
SYSTEM" was submitted to the Bharathidasan University, in partial ful fillment of the
requirements for the award of the Degree of the Bachelor of Science in Computer Science is a
record of original research work done by me during the academic year 2023-2026 under the
supervision and guidance and it has not formed the award of any Degree / Diploma
/Associateship / Fellow ship of other similar title to any candidate of any University.

SIGNATURE OF THE HOD SIGNATURE OF THE GUIDE

INTERNAL EXAMINER EXTERNAL EXAMINER

DATE:

PLACE:
DECLARATION
 DECLARATION
[Link] hereby declare that the dissertation entitled "VOICE TO
VISUAL STORY GENERATOR SYSTEM" was submitted to the
Bharathidasan University, in partial fulfilment of the requirements for the award
of the Degree of the Bachelor of Science in Computer Science is a record of
original research work done by me during 2025-2026 under the supervision and
guidance of Mrs E. Sumathi BCA., [Link]., [Link] Assistant professor,
Department of Computer Science Karur Velalar College of Arts and Science
for Women, Karur-639 111 and it has not formed the basis for the award of any
Degree/Diploma/Associate ship/ Fellowship of other similar title to any candidate
of any University.

DATE: SIGNATURE OF THE CANDIDATE

PLACE:
ACKNOWLEDGEMENT
 ACKNOWLEDGEMENT
If words are considered as symbol of approval and tokens of acknowledgement then let
the words play vital roles as expressing my gratitude.

I thank the Almighty God, the guiding light of life for granting me the strength and courage to
complete this thesis work.

I express my profound heartfelt gratitude to our Principal DR. M. MANO SAMUEL

MBA, Ph.D., [Link]., Psy.D Karur Velalar College of Arts and Science for Women Karur-639
111, for his moral support and providing me with all the facilities during the course of study.

I am extremely grateful to express my sincere thanks to my Guide Mrs E. Sumathi BCA.,

[Link]., [Link]. Assistant professor, and Head of the Department Mrs. A. KARTHIKA
,Karur Velalar College of Arts and Science for Women, Karur-639111, for the constant
support, encouragement, valuable guidance and for the help rendered to me during the course
of the thesis work.

I convey my sincere thanks to all the Staff Members and my dear friends who directly or
indirectly have given valuable guidance and support throughout the course of the thesis work.

I would like to express my heartfelt gratitude to my parents for their constant

encouragement, love, and support throughout the successful completion of this project. I
sincerely thank my friends for their valuable help, motivation, and cooperation during every
stage of this project work.
 TABLE OF CONTENTS
[Link]
TITLE [Link].

1. INTRODUCTION
1

2. LITERATURE REVIEW 3

3 RELATED WORK 5

4 PROPOSED WORK 14

5 PROPOSED RESEARCH MODEL 24

6 PERFORMANCE EVALUATION 27

7 RESULT ANALYSIS 31

8 METHODOLOGY 35

9 FUTURE ENHANCEMENT 37

10 BIBLIOGRAPHY 38

11 DISCUSSIONS 39

12 CONCLUSION 41

13 APPENDIX 42

13.1 SAMPLE CODE

13.2 SAMPLE OUTPUT
ABSTRACT
 ABSTRACT
The Voice to visual Story Generator System in PHP is a web-based application designed
to automatically generate creative and meaningful stories based on user input. The system
allows users to select genres, themes, characters, and settings, and then dynamically produces
unique stories using predefined templates and logical algorithms. Traditional story writing
requires time, creativity, and effort, which may be challenging for beginners. This system
simplifies the process by using PHP for server-side scripting and a database to store story
elements, ensuring dynamic content generation.

The application provides a user-friendly interface where users can easily enter inputs and
instantly receive generated stories. It can be used for educational purposes, creative writing
practice, entertainment, and content development. The system demonstrates the effective use
of PHP, database management, and web technologies in building an intelligent text-generation
platform.
INTRODUCTION
LITERATURE REVIEW
RELATED WORK
PROPOSED WORK
PROPOSED RESEARCH
MODEL
PERFORMANCE
EVALUATION
RESULT ANALYSIS
METHODOLOGY
FUTURE ENHANCEMENT
REFERENCES
DISCUSSION
CONCLUSION
 [Link]
In the modern digital era, technology has transformed the way people create, share, and
consume content. One of the most significant developments in recent years is the automation
of creative processes through software applications. Among these innovations, story generation
systems have gained attention for their ability to produce structured and meaningful narratives
automatically. The Voice to Visual Story Generator System in PHP is a web-based
application designed to generate creative stories dynamically based on user input. It combines
programming logic, database management, and web technologies to assist users in generating
unique and engaging stories instantly.

Storytelling has always been an essential part of human culture. From ancient oral
traditions to printed books and digital media, stories have played a crucial role in education,
entertainment, communication, and cultural preservation. However, writing a compelling story
requires imagination, creativity, vocabulary skills, and time. Many students, beginners, or
content creators may struggle with generating ideas or structuring their thoughts effectively.
The Story Generator System addresses this challenge by providing a digital platform that
simplifies the process of story creation.

The primary purpose of the Story Generator System is to automate the generation of
stories using predefined templates, stored data, and logical algorithms. The system allows users
to select various story elements such as genre (adventure, horror, romance, science fiction,
fantasy, etc.), characters, settings, and themes. Based on these inputs, the application
dynamically constructs a coherent narrative by inserting appropriate content into structured
templates. This ensures that each generated story is unique while maintaining logical flow and
readability. The system is developed using PHP as the core server-side scripting language.
dynamically PHP is widely used for web development because of its simplicity, flexibility, and
compatibility with various database systems such as MySQL. By using PHP, the application
processes user inputs, retrieves relevant data from the database, and generates story content

Problem Definition

In today’s digital environment, the rapid growth of image data has created a need for
efficient and automated image recognition systems. Manual image classification is time-
consuming, error-prone, and inefficient when dealing with large datasets. Organizations and

1
individuals often struggle to accurately analyze and categorize images due to the complexity
of visual patterns and variations in image quality, size, lighting, and background.

Traditional systems lack scalability, speed, and adaptability to dynamic data. Moreover,
inconsistent preprocessing methods and unstructured datasets reduce classification accuracy.
reliability and efficiency.

Scope of the Project

The scope of the project focuses on designing, developing, and deploying an intelligent
image classification system that can handle structured image datasets efficiently. The project
includes data collection, preprocessing (resizing and pixel normalization), feature extraction,
model training, evaluation, and cloud deployment.

The system is designed to:

 Automatically classify images into predefined categories using machine learning

algorithms.
 Provide a user-friendly web interface for uploading images and viewing results.
 Integrate backend services and cloud infrastructure for scalable performance.
 Generate performance reports using evaluation metrics such as accuracy, precision,
recall, and F1-score.

The project can be applied in various domains such as healthcare imaging, product quality
inspection, security surveillance, agriculture monitoring, and smart systems.

However, the scope is currently limited to image-based classification within selected

categories. Future improvements may include real-time video processing, advanced deep
learning architectures, and integration with IoT devices for broader real-world applications.

2
 [Link] REVIEW
[Link] and Gupta (2018):

Kumar and Gupta (2018) explored the application of deep residual networks (ResNets) for
logo verification tasks. Their research highlighted ResNets' ability to learn deep representations
and effectively mitigate gradient vanishing issues encountered in deeper networks.

Comparative evaluations with other CNN architectures demonstrated that ResNets

consistently delivered superior performance in terms of accuracy and reliability for fake logo
detection.

This study underscored the advantages of leveraging state-of-the-art deep learning

architectures to achieve robust and scalable solutions in logo authentication systems.

2. Nguyen et al. (2020):

Nguyen et al. (2020) proposed a hybrid approach combining deep learning with traditional
image processing techniques for counterfeit logo detection.

Their method integrated edge detection and colour histogram analysis as pre-processing
steps before CNN classification.

By leveraging domain-specific knowledge and deep learning capabilities, the hybrid

approach aimed to enhance model robustness against complex backgrounds and image
distortions commonly found in counterfeit logos.

The study demonstrated improved detection performance compared to purely CNN-based

methods, showcasing the synergy between deep learning and traditional image analysis
techniques.

3. Chen et al. (2022):

Chen et al. (2022) introduced a novel approach using generative adversarial networks
(GANs) to generate synthetic counterfeit logos for training deep learning models.

By creating a diverse and realistic dataset of fake logos, the authors aimed to improve model
generalization and resilience against unseen counterfeits.

3
 Their research highlighted the effectiveness of GANs in augmenting training data and
demonstrated significant advancements in detection accuracy compared to conventional data
augmentation methods.

This study exemplified innovative strategies for leveraging synthetic data to enhance deep
learning models' performance in detecting counterfeit logos.

4
 [Link] STUDY

3.1. HARDWARE REQUIREMENTS

● Processor: Intel Core i3 or higher

● RAM: Minimum 4 GB (8 GB recommended)

● Hard Disk: Minimum 250 GB

● Internet Connection (for web-based access)

● Operating System: Windows 10 / Windows 11 / Linux

3.2. SOFTWARE REQUIREMENTS

● Frontend Technologies: HTML, CSS, JavaScript

● Backend Technologies: PHP

● Database: MySQL

● Web Server: Apache / Tomcat

● Tools: Visual Studio Code

SOFTWARE DESCRIPTION

Operating System

The system can run on Windows, Linux, or macOS platforms. The operating system
provides the environment for executing the web server, database server, and application files.
It ensures process management, security, and hardware interaction.

5
HTML (Hypertext Markup Language)

HTML is used to design the basic structure of the web pages. It creates input forms for
entering certificate details such as participant name, course name, event title, date, and
certificate type. HTML ensures proper organization of content and easy navigation.

CSS (Cascading Style Sheets)

CSS is used to enhance the appearance of the application. It controls layout, colors, fonts,
spacing, and overall visual design. CSS helps maintain a uniform and professional look for all
pages and certificates

6
SOFTWARE DESCRIPTION

2.1 FRONT-END

PHP

PHP (Hypertext Processor) is a general-purpose scripting language and interpreter that

is freely available and widely used for web development. The language is used primarily for
server-side scripting, although it can also be used for command-line scripting and, to a limited
degree, desktop applications. The acronym PHP was originally derived from Personal Home
Page Tools, but it now stands for PHP: Hypertext Preprocessor, which the PHP Group's
documentation describes as a "recursive acronym."

When used for server-side scripting, PHP is added to a webpage for the purpose of
generating dynamic content when the page is accessed through a client browser. The web
server runs the script before transmitting the page to the browser. To support this process, the
web server requires PHP to be installed on the server, along with a PHP parser either a
Common Gateway Interface (CGI) parser or a server module.

When a user requests a webpage from the server, the parser interprets the PHP portion
of the page, performs the operations called for in the PHP script, and generates the Hypertext
Markup Language (HTML) that results from those operations. The HTML is then sent to the
client browser, along with any other HTML on the page, providing a seamless rendering of the
content. WebPages that contain PHP script are considered to be dynamic HTML pages because
content varies based on the results of interpreting the script.

PHP is commonly used to generate dynamic content on web pages, which is essential
for systems where user interaction and real-time updates are required. It is free and open-
source, meaning we can download it for free, and anyone can use it, modify it, and distribute.
This encourages a sizable and vibrant developer community that uses forums, tutorials, and
documentation to support and contribute to its development It is a reliable language due to its
efficient memory management and session management. It avoids unnecessary memory
[Link] code runs in its own memory space which makes it faster compared to other
scripting languages making it more efficient. In PHP, the database connections are also fast .

7
FEATURES OF PHP

PHP has incorporated numerous features. It is being consistently upgraded with new
features and code revisions. In this chapter, let's highlight some of the key features of PHP:

1. PHP is Simple and Easy to Learn

The syntax of PHP compared to that of C, Java, and Perl, which makes it rather simple
for developers to comprehend, particularly for those who are already familiar with other
programming languages. Web apps can be developed quickly because of its generous pre-
defined functions.

2. PHP is Open Source

PHP is free and open-source, meaning we can download it for free, and anyone can use
it, modify it, and distribute. This encourages a sizable and vibrant developer community that
uses forums, tutorials, and documentation to support and contribute to its development.

3. PHP is Cross-Platform Compatible

Numerous operating systems including Windows, Linux, macOS, and UNIX; and
different databases like MongoDB, PostgreSQL, MySQL are compatible with PHP.

PHP-based apps can operate on several environments without requiring any

modifications due to this cross-platform inter-operability.

8
4. Server-Side Scripting in PHP

PHP is mainly used for server-side scripting, which runs scripts on the web server and
then forwards the HTML they process to the web browser on the client. It helps the developers
in Form Submission and Session Management with users across multiple requests.

5. PHP Supports Easy Integration with Databases

PHP offers strong database interaction support for various DBMS. It offers numerous
built-in functions to achieve the database connection.

PHP also includes database abstraction layer which integrates the communication
between the application and the database. This makes it simple for developers to design
database-driven web applications.

6. PHP Provides Extensive Library Support

PHP provides extensive libraries for various functionalities like image processing,
encryption, PDF generation, parsing XML and JSON, handling sessions and cookies, and much
more.

7. Security Features in PHP

PHP provides a plethora of built-in functions for data encryption. Developers can also
leverage third-party applications for security.

PHP employs security algorithms like Sha1 and MD5 to encrypt strings. Additionally,
functions like filter_var and strip_tags contribute in maintaining a secure environment for the
users. PHP also supports secure communication protocols like HTTPS .

8. Efficient Memory and Session Management in PHP

PHP is a reliable language due to its efficient memory management and session
management. It avoids unnecessary memory [Link] code runs in its own memory
space which makes it faster compared to other scripting languages making it more efficient. In
PHP, the database connections are also fast.

9
9. PHP Has Active Community and Support

Since PHP is an open-source platform, it has a vibrant community of developers who

actively contribute to its development, share knowledge, provide support, and create third-party
tools and frameworks. Due to this active community support, PHP remains up-to-date and
developers can easily seek help from other community members in case they get any errors or
exceptions while writing PHP codes.

10. PHP has Webservers support

PHP is compatible with almost all local servers used today like Apache, Netscape,
Microsoft, IIS,etc. It is secure language to develop the website. It consists of multiple layers of
security to prevent threads and malicious attack. It is very easy to create a simple PHP example.
To do so , create a file and write html tags and PHP code and save this file with .php extension

10
2 .2 BACK-END

MySQL

MySQL is an open-source relational database management system (RDBMS) that’s

used to store and manage data. Its reliability, performance, scalability, and ease of use make
MySQL a popular choice for developers.

Databases are the essential data repositories for all software applications. For example,
whenever someone conducts a web search, logs into an account, or completes a transaction, a
database stores the information so it can be accessed in the future. MySQL excels at this
[Link], which stands for Structured Query Language, is a programming language that’s used
to retrieve, update, delete, and otherwise manipulate data in relational databases. MySQL is
officially pronounced “My ess-cue-el,” but “my sequel” is a common variation. As the name
suggests, MySQL is a SQL-based relational database designed to store and manage structured
data. In recent years, however, Oracle added additional support, including for the popular JSON
data type.
MySQL is an open-source RDBMS that uses SQL to create and manage databases. As
a relational database, MySQL stores data in tables of rows and columns organized into
schemas. A schema defines how data is organized and stored and describes the relationship
among various tables. With this format, developers can easily store, retrieve, and analyze many
data types, including simple text, numbers, dates, times, and, more recently, JSON and vectors.

Benefits of MySQL

MySQL is fast, reliable, scalable, and easy to use. It was originally developed to handle
large databases quickly and has been used in highly demanding production environments for
many years. MySQL offers a rich and useful set of functions, and it’s under constant
development by Oracle, so it keeps up with new technological and business demands.
MySQL’s connectivity, speed, and security make it highly suited for accessing databases on
the internet. It is one of the most mature and widely used databases. It has been tested in a wide
variety of scenarios for nearly 30 years, including by many of the world’s largest companies.
For business-critical applications and service level agreement commitments, customers can
achieve recovery point objective zero (zero data loss) and recovery time objective zero seconds.

11
MySQL’s key benefits include the following:-

[Link] of use:
Developers can install MySQL in minutes, and the database is easy to
manage.

[Link]:
MySQL is one of the most mature and widely used databases. It has been
tested in a wide variety of scenarios for nearly 30 years, including by many of the
world’s largest companies. Organizations depend on MySQL to run business-critical
applications because of its reliability.
3. Scalability:
MySQL scales to meet the demands of the most accessed applications.
MySQL’s native replication architecture enables organizations, including Facebook,
Netflix, and Uber, to scale applications to support tens of millions of users or more.
4. Performance:
MySQL is a proven high performance, zero-administration database system
and comes in a range of editions to meet nearly any demand. Cloud-based Heatwave
MySQL provides unmatched performance and price-performance, according to
industry benchmarks including TPC-H, TPC-DS, and CH-benCHmark.
5. High availability:
MySQL delivers a complete set of native, fully integrated replication
technologies for high availability and disaster recovery. For business-critical
applications and service level agreement commitments, customers can achieve
recovery point objective zero (zero data loss) and recovery time objective zero
seconds (automatic failover).
6. Security:
Data security entails both data protection and compliance with industry and
government regulations, including the European Union General Data Protection
Regulation, the Payment Card Industry Data Security Standard, the Health Insurance
Portability and Accountability Act, and the Defense Information Systems Agency’s
Security Technical Implementation Guides. MySQL Enterprise Edition provides
advanced security features, including authentication/authorization, transparent data
encryption, auditing, data masking, and a database firewall.

12
7. Flexibility:
The MySQL Document Store gives users maximum flexibility in developing
traditional SQL and NoSQL schema-free database applications. Developers can mix
and match relational data and JSON documents in the same database and application
8. Cost-Effective:
MySQL is open-source software, which means it is free to use and modify.
This makes it an affordable choice for businesses, especially small and medium-
sized enterprises.
9. Community Support:
Being open-source, MySQL has a large, active community that provides
free resources, documentation, and support.
[Link] Data Processing:
MySQL is known for its fast read and write operations, making it suitable
for high-traffic websites and applications.
[Link]:
MySQL supports a variety of indexing techniques, which improve query
performance by enabling faster data retrieval.

13
 [Link] WORK

Early Rule-Based Narrative Systems

Early story generation systems relied heavily on predefined rules and symbolic logic. One
foundational example is TALE-SPIN, developed by James R. Meehan in the 1970s. TALE-
SPIN generated simple fables by simulating character goals and problem-solving behaviors.

Similarly, UNIVERSE focused on character-driven simulations, where events unfolded

based on predefined world models. These systems emphasized logical consistency but often
lacked linguistic richness and emotional depth. Their contribution lies in establishing
foundational concepts such as goal-driven agents and causal reasoning in narratives.

Planning-Based Narrative Generation

Planning-based systems introduced AI planning algorithms to create structured and goal-

oriented plots. A prominent example is IPOCL, developed by Mark O. Riedl and colleagues.
IPOCL improved narrative coherence by ensuring that character actions aligned with
believable intentions.

Another influential framework is Façade, created by Michael Mateas and Andrew Stern.
Façade blended AI planning with interactive storytelling, allowing users to influence narrative
progression in real time. These systems bridged the gap between static story output and
dynamic narrative interaction.

Machine Learning and Neural Language Models

With the rise of machine learning, story generation shifted toward data-driven approaches.
Neural language models, particularly transformer architectures, significantly improved fluency
and contextual understanding.

Models such as GPT-3 by OpenAI demonstrated the ability to generate long-form, coherent
narratives from minimal prompts. Similarly, BERT by Google advanced contextual text
representation, influencing narrative generation research.

These neural systems excel in linguistic creativity and adaptability but sometimes struggle with
long-term plot consistency compared to symbolic planners.

14
Interactive and Game-Based Storytelling

In gaming environments, procedural narrative generation has played a crucial role. Games
like AI Dungeon utilize large-scale language models to create open-ended interactive stories.
This approach allows users to shape narratives dynamically, enhancing immersion and
replayability.

Research in interactive storytelling also explores adaptive plot systems that respond to player
decisions, combining reinforcement learning with narrative modeling. These systems prioritize
user agency and experiential storytelling.

Hybrid and Multimodal Approaches

Recent studies explore hybrid models that integrate symbolic reasoning with neural
networks. These systems aim to combine structured plot planning with expressive language
generation. Multimodal story generation—incorporating text, images, and audio—is also
emerging, expanding narrative immersion.

For example, research on visual storytelling links image recognition models with narrative
text generation, enabling coherent storylines based on visual sequences.

15
PROPOSED WORK

The proposed Story Generator System is a web-based application developed using PHP and
MySQL to automatically generate structured and meaningful stories based on user input. The
system is designed to provide an interactive platform where users can select story elements
such as genre, characters, setting, theme, and plot type. Based on these inputs, the system
dynamically generates a complete story using predefined templates and logical algorithms. The
primary objective of the proposed system is to simplify the creative writing process while
maintaining narrative coherence and originality.

The architecture of the system follows a client-server model. The frontend is developed
using HTML, CSS, and basic JavaScript to ensure a user-friendly interface. The backend is
implemented using PHP, which processes user inputs, retrieves relevant data from the database,
and constructs the final story output. The system ensures proper validation of inputs and
efficient handling of database queries to maintain performance and reliability.

Data Collection

Data collection plays a crucial role in the development of the Story Generator System. The
system requires a well-structured database containing story elements such as character names,
roles (hero, villain, mentor, etc.), locations, events, conflicts, dialogues, and endings. These
data elements are collected from various sources including storybooks, online public-domain
stories, educational materials, and creative writing samples. The collected content is
categorized according to genres such as adventure, horror, romance, fantasy, and science
fiction.

Each story component is stored in separate database tables to allow flexible and dynamic
combination during story generation. The data is carefully reviewed and organized to ensure
grammatical correctness and logical consistency. Randomization techniques are applied while
selecting elements to ensure uniqueness in generated stories.

By combining structured data collection with dynamic PHP-based processing, the proposed
system ensures efficient story generation while maintaining creativity, coherence, and scalabil

16
[Link] Expected Outcome Description
1 Automated Story The system will automatically generate complete and
Generation meaningful stories based on user-selected inputs such as
genre, characters, and setting.

2 User-Friendly Users will be able to easily select story elements and generate
Interface stories through a simple and interactive web interface.

3 Dynamic Stories will be generated dynamically using PHP and

Content Creation database-driven templates, ensuring variation in each output.

4 Reduced The system will minimize manual effort required to create

Writing Effort stories, helping beginners overcome writer’s block.

5 Structured Generated stories will follow a proper structure including

Story Format introduction, conflict, climax, and conclusion.

6 Multiple Genre The system will support various genres such as adventure,
Support horror, romance, fantasy, and science fiction.

7 Database Story elements will be stored and managed efficiently in a

Management structured MySQL database.
Efficiency

8 Fast Processing Stories will be generated instantly with minimal processing

Time delay.

9 Educational Support The system will help students understand story construction
and improve creative writing skills.

17
Expected Result:
1. Data Preprocessing:

Data preprocessing is a crucial step to ensure the accuracy and efficiency of the system. In
the proposed model, raw input data such as images or text inputs are cleaned, filtered, and
organized before being used for training or processing. The preprocessing stage removes
noise, irrelevant information, and inconsistencies that may affect the system’s performance.
For image-based inputs, unwanted background variations, distortions, and low-quality
samples are handled during this phase.

The dataset is carefully examined to ensure that it contains balanced samples for each
category. Duplicate entries are removed to prevent biased learning. Missing or corrupted files
are identified and either corrected or excluded from the dataset. Additionally, data labeling is
verified to ensure that each input is correctly assigned to its respective class.

By performing preprocessing, the system improves model reliability, reduces

computational errors, and enhances overall performance. Clean and structured data allows the
classification model to learn meaningful patterns effectively, leading to higher prediction
accuracy and better generalization on unseen data.

2. Resizing Images

Resizing images is an essential preprocessing step to maintain uniformity in the dataset.

Since images may have different dimensions and resolutions, they are resized to a fixed size
before being fed into the model. This ensures consistency in input shape and reduces
computational complexity.

By standardizing image dimensions, the system can process data efficiently without
memory overflow issues. Resizing also helps in faster training and testing, as smaller image
sizes reduce processing time. Proper resizing preserves important visual features while
ensuring compatibility with the classification model.

18
3. Pixel Normalisation

Pixel normalization is performed to scale image pixel values into a standard range,
typically between 0 and 1. Normally, image pixel values range from 0 to 255. Directly using
these large values can slow down the training process and reduce model efficiency.
Therefore, normalization improves numerical stability and enhances convergence speed
during model training.

By dividing each pixel value by 255, the dataset is transformed into a smaller and
consistent range. This ensures that no single feature dominates others due to higher
magnitude. Pixel normalization also helps the model detect patterns more effectively by
balancing brightness and contrast levels across images.

Another important benefit of normalization is improved gradient flow during

backpropagation. Large input values may cause unstable gradients, leading to slower
convergence or vanishing/exploding gradient problems. Normalized inputs ensure smoother
and faster learning.

Furthermore, normalization increases model generalization by reducing sensitivity to

lighting variations and minor intensity changes. This allows the system to perform better on
new, unseen data.

Key Points of Pixel Normalization:

1. Scales pixel values from 0–255 to 0–1 range.

2. Improves training speed and convergence rate.
3. Prevents dominance of high-value features.
4. Enhances numerical stability in computations.
5. Increases model accuracy and generalization ability.

4. Classification

Classification is the final stage where the processed data is categorized into predefined
classes. After preprocessing, resizing, and normalization, the model analyzes extracted
features to determine the correct class label. The classification algorithm learns patterns from
training data and applies them to predict outcomes for new inputs.

19
 The expected result is accurate and reliable classification with minimal error rates. The
system aims to achieve high precision and recall by correctly identifying categories. Proper
preprocessing and normalization significantly improve classification performance and ensure
consistent results across different datasets.

Components of Proposed System Diagram

1. User Interface (UI)

The User Interface is the front-end component that allows users to interact with the
system. It provides forms, dashboards, and visual elements for input and output display. The
UI ensures smooth navigation and easy access to system features. It is designed to be
responsive, interactive, and user-friendly.

 Enables users to upload images and view results.

 Provides interactive dashboards and notifications.

2. Back-End Services

Back-end services handle the core processing logic of the system. They manage data
requests, authentication, and communication between modules. This layer processes user
inputs and coordinates with databases and machine learning models. It ensures secure and
efficient system operations.

 Handles API requests and server-side processing.

 Manages database connectivity and security.

3. Cloud Infrastructure

Cloud infrastructure provides scalable storage and computing resources for the system. It
supports data hosting, processing power, and deployment services. The cloud ensures high
availability and reliability of the application. It allows flexible expansion based on user
demand.

 Stores datasets and model files securely.

 Enables scalable computing and remote access.

20
4. Image Recognition Module

The Image Recognition Module processes uploaded images to extract meaningful features.
It uses preprocessing techniques and trained models to identify patterns. This module plays a
key role in detecting and classifying images accurately. It ensures fast and reliable image
analysis.

 Performs image preprocessing and feature extraction.

 Detects and classifies images into categories.

5. Recommendation Engine
The Recommendation Engine suggests relevant outputs based on user behavior and
classification results. It analyzes patterns and preferences to generate personalized
recommendations. This component enhances user experience by providing intelligent
suggestions. It adapts over time using learning mechanisms.

 Provides personalized suggestions to users.

 Improves engagement and satisfaction.

6. Data Analytics and Reporting

This component analyzes system data to generate meaningful insights. It tracks performance
metrics, usage statistics, and classification accuracy. Reports are generated for monitoring
and decision-making purposes. It helps in evaluating and improving system efficiency.

 Generates performance and usage reports.

 Supports data-driven decision making.

7. Location Services
Location Services identify and utilize geographical information for enhanced functionality.
It helps in tagging and tracking image data based on location. This component improves

21
contextual accuracy and personalization. It integrates GPS or map-based services when
required.

 Provides location-based data tagging.

 Enhances contextual analysis and tracking.

8. Machine Learning Model Management

Machine Learning Model Management handles model training, validation, deployment,
and updates. It ensures that models are optimized and maintained efficiently. This component
monitors performance and retrains models when necessary. It guarantees consistent and
accurate predictions over time.

 Manages model training and deployment lifecycle.

 Monitors performance and updates models regularly.

22
 5. PROPOSED RESEARCH MODEL

The proposed research module focuses on designing and implementing an intelligent image-
based recognition and classification system integrated with modern web technologies and
machine learning techniques. The module aims to develop a scalable, accurate, and efficient
framework that can process image data, extract meaningful features, classify inputs, and
provide analytical insights. The research emphasizes structured data handling, preprocessing
techniques, model optimization, and performance evaluation to ensure reliable results. The
system integrates preprocessing pipelines, machine learning algorithms, cloud deployment, and
reporting mechanisms to create a comprehensive solution. The following subtopics describe
the major components of the proposed research module.

1. Data Acquisition and Dataset Preparation

Data acquisition is the first stage of the research module. The dataset consists of labeled
images collected from reliable sources such as open datasets, publicly available repositories,
and manually gathered samples. The collected data is categorized into predefined classes to
ensure balanced representation.

After collection, the dataset undergoes validation to remove duplicate, corrupted, or low-
quality images. Proper labelling is verified to prevent misclassification. The dataset is then
divided into training, validation, and testing subsets. This structured preparation ensures
unbiased model evaluation and improves generalization capability. Data augmentation
techniques such as rotation, flipping, and zooming are applied to increase dataset diversity and
reduce overfitting.

2. Data Preprocessing and Feature Engineering

Data preprocessing ensures that raw images are transformed into a suitable format for model
training. Images are resized into uniform dimensions to maintain consistency across the dataset.
Pixel normalization is applied to scale pixel values into a smaller range, typically between 0
and 1, to improve numerical stability.

Feature engineering techniques are used to extract meaningful attributes from images. These
features may include edges, textures, color histograms, and shape descriptors. In deep learning
approaches, convolutional layers automatically extract hierarchical features. Proper

23
preprocessing enhances model efficiency, accelerates training, and improves overall
classification accuracy.

3. Model Development and Training

The model development phase involves selecting appropriate machine learning or deep
learning algorithms for classification. Convolutional Neural Networks (CNNs) are commonly
used for image recognition due to their ability to capture spatial hierarchies in images. The
model architecture is carefully designed with convolutional layers, pooling layers, and fully
connected layers.

Training is performed using the prepared dataset. Hyperparameters such as learning rate,
batch size, and number of epochs are tuned for optimal performance. Loss functions and
optimization algorithms are selected to minimize classification errors. Regularization
techniques like dropout are applied to prevent overfitting. The training process is monitored
using validation metrics to ensure consistent improvement.

4. System Integration and Cloud Deployment

After model training, the system is integrated into a web-based platform. The backend
services manage communication between the user interface and the machine learning module.
The trained model is deployed on cloud infrastructure to enable scalability and remote access.

Cloud deployment ensures efficient resource utilization and allows real-time image
processing. APIs are developed to handle image uploads, model inference, and result delivery.
Security measures such as authentication and encrypted data transmission are implemented to
protect user data. This integration ensures smooth system functionality and accessibility.

5. Performance Evaluation and Reporting

Performance evaluation is a critical stage in the research module. The trained model is tested
using unseen test data to measure accuracy, precision, recall, and F1-score. Confusion matrices
are generated to analyse classification results in detail.

Data analytics tools are integrated to monitor system usage and prediction performance.
Reports are generated to provide insights into model behaviour and accuracy trends.

24
 [Link] EVALUATION

Performance evaluation is conducted to measure the effectiveness, accuracy, and reliability

of the proposed classification system. It helps in understanding how well the model performs
on unseen data and whether it meets the expected objectives. Evaluation metrics provide
quantitative measures that indicate strengths and weaknesses of the system. By analyzing these
metrics, improvements can be made to optimize performance and reduce errors. The following
evaluation metrics are used to assess the system.

Evaluation Metrics

1. Accuracy

Accuracy measures the overall correctness of the model by calculating the ratio of correctly
predicted instances to the total number of predictions. It indicates how often the classification
model provides the right output. While accuracy is simple and widely used, it may not be
sufficient when dealing with imbalanced datasets.

Formula:
Accuracy = (TP + TN) / (TP + TN + FP + FN)

2. Precision

Precision measures the proportion of correctly predicted positive instances out of all
predicted positive instances. It focuses on the quality of positive predictions made by the
model. High precision indicates that the system produces fewer false positives and is reliable
in its positive classifications.

Formula:
Precision = TP / (TP + FP)

26
3. Recall (Sensitivity)

Recall measures the proportion of correctly predicted positive instances out of all actual
positive instances. It evaluates the model’s ability to identify all relevant samples in the dataset.
High recall means the system successfully detects most of the true positive cases with fewer
false negatives.

Formula:
Recall = TP / (TP + FN)

4. F1-Score

F1-Score is the harmonic mean of precision and recall. It provides a balanced evaluation
metric when both false positives and false negatives are important. This metric is particularly
useful when the dataset is imbalanced, as it considers both precision and recall together.

Formula:
F1-Score = 2 × (Precision × Recall) / (Precision + Recall)

5. Confusion Matrix

A confusion matrix is a tabular representation of actual versus predicted classifications. It

displays four key components: True Positives (TP), True Negatives (TN), False Positives (FP),
and False Negatives (FN). The confusion matrix provides a detailed breakdown of
classification performance and helps in identifying specific types of errors made by the model.

27
IMPLEMENTATION

The implementation phase is the stage where the system design is converted into a fully
functional Data Analytics. It his phase, all modules are developed, integrated, tested, and
prepared for actual use.

1. Environment Preparation

● Required hardware and software are installed.

● Microsoft Excel is configured as the database.

● Necessary development tools are set up.

2. Database Creation

● Excel worksheets are created for products, suppliers, users, stock-in, stock-out, and
reports.

● Proper column structures and unique IDs are defined.

● Data validation rules are applied.

3. Module Development

● User Management Module is implemented for secure authentication and role-based

access control.

● Admin Module is developed to process online contributions and maintain user

records.

● Register Module is implemented to register users and manage activity assignments.

● Story Generator Module is developed to monitor project progress and real-time fund
utilization.

4. Analytics Integration

● Pivot tables and formulas are used to analyse data.

● Fast-moving and slow-moving item analysis is implemented.

● Reorder level alerts are generated.

28
5. Report Implementation

● Automated reports are designed for daily, monthly, and yearly summaries.

● Charts and graphs are added for visual representation.

6. Data Security and Backup

● Sheet protection and password security are applied.

● Regular backup mechanisms are implemented to prevent data loss.

7. Testing and Validation

● Each module is tested individually.

● System testing is performed using sample data.

● Errors are identified and corrected.

8. Deployment

● The system is deployed for real-time usage.

● Users are trained to operate the system effectively.

29
 [Link] ANALYSIS
The result analysis of the proposed image recognition and classification system
demonstrates its effectiveness, reliability, and practical applicability. After implementing data
preprocessing techniques such as resizing and pixel normalization, the model was trained and
evaluated using structured datasets. The analysis focuses on performance metrics, prediction
consistency, error patterns, and overall system efficiency.

The trained model achieved high classification accuracy on the testing dataset, indicating
that the preprocessing and feature extraction methods significantly improved learning
capability. The confusion matrix analysis showed that most samples were correctly classified,
with minimal false positives and false negatives. Precision values were high, demonstrating
that the model rarely misclassified negative samples as positive. Similarly, recall scores
indicated that the system successfully detected the majority of actual positive instances.

The F1-score provided a balanced evaluation, confirming that the model maintains
consistency between precision and recall. This balance ensures reliable predictions even when
class distributions vary. Performance comparisons between training and validation datasets
showed minimal overfitting, proving that the model generalizes well to unseen data.

In addition to accuracy, computational efficiency was analyzed. The preprocessing pipeline

reduced noise and unnecessary variations, leading to faster training convergence. Image
resizing decreased memory usage and improved processing speed, while pixel normalization
enhanced gradient stability and model optimization.

Cloud deployment testing confirmed that the system can handle multiple image uploads
simultaneously without significant latency. The integration between the front-end interface and
backend services operated smoothly, providing real-time prediction results. Reporting and
analytics modules generated detailed performance summaries for monitoring and further
improvements.

Overall, the result analysis indicates that the proposed system achieves strong classification
performance, maintains stability during deployment, and delivers reliable outputs. Minor errors
observed were primarily due to visually similar image classes, suggesting that future
improvements can include larger datasets and advanced feature extraction techniques. The

30
system meets its intended objectives and demonstrates scalability, efficiency, and accuracy in
practical scenarios. A Story Generator is an intelligent narrative system designed to transform
prompts, themes, or keywords into fully developed stories. It blends creativity with algorithmic
structure to produce coherent plots, engaging characters, and immersive settings. The result
analysis of such a system evaluates how effectively it meets storytelling standards while
maintaining originality, coherence, and emotional depth. This analysis explores its
performance across creativity, structure, engagement, adaptability, and overall impact.

Creativity and Originality

One of the most significant strengths of the Story Generator lies in its imaginative capacity.
The system demonstrates an ability to produce unique plotlines, unexpected twists, and diverse
genres—from fantasy epics to psychological thrillers. Unlike traditional templates, it adapts
dynamically to user prompts, ensuring that no two stories feel identical.

The generator excels in combining familiar tropes with innovative variations. For example,
a classic “hero’s journey” narrative may be enriched with unconventional character motivations
or morally complex conflicts. This balance between familiarity and novelty enhances reader
satisfaction and prevents predictability. The creativity output reflects a strong internal narrative
model capable of synthesizing ideas across themes and tones.

Plot Structure and Coherence

A compelling story depends on logical progression, and the Story Generator performs
effectively in maintaining structured storytelling. It typically follows a clear arc—introduction,
rising action, climax, falling action, and resolution—ensuring narrative completeness.

Transitions between scenes are generally smooth, with consistent cause-and-effect

relationships. Conflicts are introduced with clarity and resolved meaningfully, avoiding abrupt
endings. The system also maintains timeline consistency and minimizes plot holes, which is
critical for reader immersion.

However, in more complex multi-layered narratives, occasional pacing imbalances may

appear, such as extended exposition or rapid resolutions. Despite this, the overall coherence
remains strong and dependable.

31
Character Development and Emotional Depth

Characterization is central to immersive storytelling. The Story Generator effectively

creates multidimensional characters with distinct personalities, motivations, and growth arcs.
Dialogue is often contextually relevant and contributes to character authenticity. Emotional
layering is another notable strength. Stories frequently evoke empathy, suspense, joy, or
tension through descriptive language and situational stakes. Internal monologues and relational
conflicts add psychological realism, making characters feel relatable rather than mechanical.

That said, deeper emotional complexity—such as subtle moral ambiguity or evolving

interpersonal dynamics—may require refined prompting for optimal results.

World-Building and Immersion

The generator demonstrates strong environmental detailing, particularly in fantasy and

science fiction genres. Settings are vividly described, incorporating sensory imagery that
enhances immersion. Whether portraying futuristic cities or enchanted forests, descriptive
elements help readers visualize the narrative world clearly.

Cultural systems, societal norms, and environmental conditions are often integrated
organically into the storyline. This immersive world-building strengthens suspension of
disbelief and contributes to narrative depth.

Adaptability and User Customization

A key performance indicator is adaptability. The Story Generator responds effectively to

varied prompts, tones, genres, and target audiences. It can produce children’s stories with
simple language or complex literary fiction with layered symbolism.

Customization options—such as specifying themes, character roles, or plot twists—enhance

user control. This flexibility makes the tool suitable for creative writers, educators, marketers,
and entertainment developers.

Overall, the Story Generator achieves a high level of narrative competence. It combines
creativity, structural integrity, emotional engagement, and adaptability into a cohesive

32
 [Link]

1. Model Architecture

Model architecture defines the structural design of the machine learning system, including
layers and computational flow. It determines how input data is processed and transformed
into output predictions.

 The architecture consists of input, convolutional, pooling, and fully connected layers
to extract and classify image features effectively.
 Activation functions such as ReLU and Softmax are used to introduce non-linearity
and generate probability-based outputs.
 Regularization techniques like dropout and batch normalization are included to
improve generalization and prevent overfitting.

2. Training Procedure

The training procedure refers to the process of teaching the model using labeled data. It
involves adjusting model parameters to minimize prediction errors.

 The dataset is divided into training and validation sets to ensure unbiased learning and
performance monitoring.
 Backpropagation and optimization algorithms such as Adam or SGD are used to
update model weights iteratively.
 Hyperparameters including learning rate, batch size, and epochs are tuned to achieve
optimal accuracy and convergence speed.

3. Evaluation Metrics

Evaluation metrics are quantitative measures used to assess model performance. They help
determine accuracy, reliability, and predictive effectiveness.

 Accuracy measures the overall percentage of correct predictions made by the model.
 Precision, recall, and F1-score evaluate classification quality, especially in handling
imbalanced data.
 A confusion matrix provides detailed insight into true positives, false positives, true
negatives, and false negatives.

34
4. Experimental Setup
Experimental setup refers to the environment and configuration used to train and test the
model. It includes hardware, software, and dataset specifications.

 The system is implemented using Python and deep learning libraries within a
structured development environment.
 Training is conducted on a system with sufficient CPU/GPU resources to handle
computational requirements.
 The dataset is preprocessed through resizing and normalization before being fed into
the model.

5. Ethical Considerations

Ethical considerations ensure that the system is developed and deployed responsibly. They
focus on fairness, privacy, and data security.

 User data and uploaded images are handled securely to maintain privacy and prevent
misuse.
 The dataset is reviewed to avoid biased or discriminatory representations in model
predictions.
 The system is designed to provide transparent results and prevent harmful or unethical
applications.

35
 [Link] ENHANCEMENTS

The proposed image recognition and classification system has significant potential for
future enhancements and expansion. As technology continues to evolve, the system can be
improved in terms of accuracy, scalability, intelligence, and real-world applicability. The
following points highlight the future scope of the project:

1. Integration of Advanced Deep Learning Models

Future versions can incorporate advanced architectures such as Efficient Net,
Vision Transformers (ViT), or hybrid models to improve feature extraction and
classification accuracy.
2. Real-Time Video Processing
The system can be extended to process live video streams instead of static
[Link] would enable real-time detection and classification for surveillance and
monitoring applications.
3. Mobile Application Development
A mobile-based version of the system can be developed to allow users to capture
and classify images directly using smartphone cameras.

36
 [Link]

REFERENCE BOOKS:
1. WEB TECHNOLOGIES – BLACK BOOK
Publisher: Dream Tech Press
2. Murach’s MySQL –3 Edition
rd

Author: Joel Murah, Publisher: Paperback

3. Programming PHP – 4 Edition
th

Author: Rusmus Learterf Publisher: O’Reilly Media

4. Beginning PHP and MySQL – 3 Edition
rd

Author: Jason Publisher: Apress

5. Sommerville, Ian. Software Engineering.
10th Edition, Pearson Education, 2016.

WEBSITE:
1. [Link]
2. [Link]
3. [Link]

37
 [Link]
A voice-to-visual story generator system is an advanced multimodal platform
that converts spoken language into dynamic visual narratives. By integrating automatic
speech recognition (ASR), natural language processing (NLP), and generative visual
models, the system transforms a user’s voice input into illustrated scenes, animations,
or cinematic sequences. This technology sits at the intersection of conversational AI
and creative media, leveraging breakthroughs similar to those seen in OpenAI’s
generative systems and image models like DALL·E.

The architecture typically consists of several interconnected modules. First, a

speech-to-text engine converts raw audio into structured text. Modern systems may rely
on transformer-based ASR models for high accuracy across accents and environments.
Next, an NLP module interprets the semantic structure of the story. It extracts
characters, settings, actions, emotions, and temporal relationships. Techniques such as
named entity recognition, sentiment analysis, and scene segmentation allow the system
to break the narrative into visualizable components.

Once the narrative structure is identified, a story planner organizes the content
into scenes. This component ensures coherence, continuity, and pacing. For instance, if
a user narrates, “A dragon flew over a castle at sunset,” the system identifies the key
elements—dragon, castle, sunset—and their relationships. A visual generation module
then produces corresponding imagery using diffusion or generative adversarial network
(GAN) models. Tools inspired by technologies like Unreal Engine can render these
scenes in real time for immersive 3D storytelling.

To maintain narrative consistency across multiple scenes, the system may use
memory networks or context-tracking algorithms. Character appearance, environmental
style, and mood must remain stable throughout the visual sequence. Advanced
implementations also incorporate style transfer, allowing users to specify artistic themes
such as watercolor, anime, or cinematic realism.

38
 From a user experience perspective, interactivity is a key strength. Users can refine
the story mid-generation, adjust visual styles, or add new plot elements verbally. This
makes the system highly valuable for education, entertainment, marketing, and
accessibility. For example, children can narrate imaginative tales and instantly see them
visualized, enhancing creativity and engagement.

However, challenges remain. Real-time rendering requires high computational

resources. Ambiguity in spoken language can lead to inaccurate visual interpretations.
Ethical considerations—such as preventing harmful or misleading visual content—must
also be addressed through content moderation frameworks.

In conclusion, a voice-to-visual story generator system represents a powerful

evolution in human-computer interaction. By combining speech understanding,
narrative intelligence, and generative visualization, it transforms storytelling into an
immersive, interactive, and highly personalized experience.

39
 [Link]

A voice-to-visual story generator system is an advanced multimodal platform

that converts spoken language into dynamic visual narratives. By integrating automatic
speech recognition (ASR), natural language processing (NLP), and generative visual
models, the system transforms a user’s voice input into illustrated scenes, animations,
or cinematic sequences. This technology sits at the intersection of conversational AI
and creative media, leveraging breakthroughs similar to those seen in OpenAI’s
generative systems and image models like DALL·E.

The architecture typically consists of several interconnected modules. First, a speech-

to-text engine converts raw audio into structured text. Modern systems may rely on
transformer-based ASR models for high accuracy across accents and environments.
Next, an NLP module interprets the semantic structure of the story. It extracts
characters, settings, actions, emotions, and temporal relationships. Techniques such as
named entity recognition, sentiment analysis, and scene segmentation allow the system
to break the narrative into visualizable components.

40
 [Link]
13.1SAMPLE CODING

INDEX:
<!DOCTYPE html>

<html lang="en">

<head>

<meta name="viewport" content="width=device-width, initial-scale=1.0" />

<title>Story Generator App</title>

<link

href="[Link]
boto+Mono:wght@400;500&display=swap"

rel="stylesheet"

/>

<link rel="stylesheet" href="./[Link]">

</head>

<body>

<button class="dark-mode-toggle" id="darkModeToggle"> </button>

<div class="container">

<h1>Story Generator App</h1>

<div class="wrapper">

<div>

<label for="name">Enter a Name:</label>

<input

type="text"

id="name"

placeholder="Type a name..."

41
 required

/>

</div>

<div>

<label for="place">Enter a Place:</label>

<input

type="text"

id="place"

placeholder="Type a place..."

required

/>

</div>

<div>

<label for="object">Enter an Object:</label>

<input

type="text"

id="object"

placeholder="Type an object..."

required

/>

</div>

<div>

<label for="action">Enter an Action:</label>

<input

type="text"

id="action"

placeholder="Type an action..."

42
 required

/>

</div>

</div>

<button id="generateStory">Generate Story</button>

<div id="story" class="story"></div>

<div class="buttons">

<button id="resetButton">Reset</button>

<button id="anotherStory">Generate Another Story</button>

</div>

</div>

<script src="./[Link]"></script>

</body>

</html>

const nameInput = [Link]("name");

const placeInput = [Link]("place");

const objectInput = [Link]("object");

const actionInput = [Link]("action");

const storyDiv = [Link]("story");

const generateButton = [Link]("generateStory");

const resetButton = [Link]("resetButton");

const anotherStoryButton = [Link]("anotherStory");

const darkModeToggle = [Link]("darkModeToggle");

const stories = [

43
 `One sunny morning, \${name} decided to visit \${place}. As they wandered around, they
stumbled upon a mysterious \${object} hidden under a tree. Curiosity got the better of them,
and they decided to \${action}. Little did they know, this simple act would change their life
forever.

As \${name} began to \${action}, the \${object} started glowing brightly. Suddenly, the
ground beneath them shook, and they were transported to a magical world filled with talking
animals and floating islands. \${name} realized they had unlocked a portal to another
dimension, and their adventure was just beginning!`,

`In the heart of \${place}, \${name} discovered an ancient \${object} buried deep in the
sand. Intrigued, they decided to \${action}, hoping to uncover its secrets. As they did, the
\${object} began to hum with energy, and a beam of light shot into the sky.

Moments later, \${name} found themselves surrounded by a group of friendly aliens. The
aliens explained that the \${object} was a key to their spaceship, and they needed \${name}'s
help to return home. Together, they embarked on an interstellar journey filled with danger
and wonder.`,

`While exploring the dense forests of \${place}, \${name} found a strange \${object} lying
on the ground. It looked out of place, so they decided to \${action}. As soon as they did, the
\${object} began to emit a soft, golden light.

The light grew brighter, and \${name} felt a strange sensation. Suddenly, they were
surrounded by a group of ancient warriors who claimed the \${object} was a relic of their
ancestors. They asked \${name} to help them protect it from an evil sorcerer, and together,
they set off on an epic quest.`,

`One stormy night, \${name} was walking through \${place} when they spotted a glowing
\${object} in the distance. Drawn to its light, they decided to \${action}. As they did, the
storm cleared, and the \${object} began to float in the air.

A voice echoed from the \${object}, revealing that it was a magical artifact with the power to
grant wishes. \${name} was given one wish, but they had to choose wisely. After much

44
thought, they made their wish, and the world around them transformed in ways they never
imagined.`,

`While hiking in the mountains of \${place}, \${name} found a peculiar \${object} half-
buried in the snow. They decided to \${action}, and as they did, the \${object} began to emit
a warm, comforting glow.

Suddenly, a portal opened, and \${name} was greeted by a wise old wizard. The wizard
explained that the \${object} was a key to a hidden kingdom, and \${name} was chosen to be
its protector. With newfound courage, they stepped through the portal and into their destiny.`,

];

function generateStory() {

const name = [Link];

const place = [Link];

const object = [Link];

const action = [Link];

if (!name || !place || !object || !action) {

alert("Please fill in all the fields!");

return;

const randomStory = stories[

[Link]([Link]() * [Link])

.replaceAll("${name}", name)

.replaceAll("${place}", place)

.replaceAll("${object}", object)

45
 .replaceAll("${action}", action);

[Link] = randomStory;

[Link] = "block";

[Link] = 1;

function resetForm() {

[Link] = "";

[Link] = "";

[Link] = "";

[Link] = "";

[Link] = "none";

[Link] = 0;

function toggleDarkMode() {

[Link]("dark-mode");

[Link]("click", generateStory);

[Link]("click", resetForm);

[Link]("click", generateStory);

[Link]("click", toggleDarkMode);

46
13.2 SAMPLE OUTPUT

47
48
49