INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 910
RideSmart: A Personalized Motorcycle Product Recommendation
System Using TF-IDF and Descriptive Analytics for Javidson
Motorshop
June Daniel Bautista
1
, Kenneth Calopez
2
, John Robert Evangelista
3
, John Michael Lorbes
4
, Enrico
Chavez
5
, Erwin Guillermo
6
1 2 3 4 5 6
(SY 2025-2026) Arellano University, Pasig Campus
DOI: https://dx.doi.org/10.51584/IJRIAS.2025.1010000074
Received: 20 September 2025; Accepted: 25 September 2025; Published: 07 November 2025
ABSTRACT
RideSmart is a web-based tool designed to provide JavidsonMotorshop clients with personalized product
recommendations, including appropriate motorcycle parts, accessories, and services. To create recommendations
that are specific to each user, the system examines purchase histories, product qualities, and consumer
preferences using TF-IDF (Term Frequency–Inverse Document Frequency) and descriptive analytics. Compared
to generic promos, this not only expedites and enhances the shopping experience but also provides insights into
consumer behavior, including demand trends, buying habits, and popular products, which aid the store in
improving its marketing and stocking plans. The RideSmart recommendation system for JavidsonMotorshop
uses TF-IDF to extract key features and keywords from customer reviews and searches, enabling personalized
motorcycle product suggestions that improve purchase decisions and customer experience (Huang, 2025).
Descriptive analytics complements this by analyzing historical sales, browsing behavior, and product interaction
data to identify trends and patterns, supporting data-driven marketing and inventory management (Harvard
Business School Online, 2021). Together, these approaches allow the system to tailor recommendations to
individual preferences, enhancing sales efficiency and operational effectiveness (Sigma Computing, 2025;
MEEGLE, 2025).Laravel, a PHP framework, Python, MySQL, and front-end technologies (HTML, CSS) are
used in the technical implementation of RideSmart. Functionality (accurate product details and seamless search,
cart, checkout, etc.), reliability (availability and recovery from errors), efficiency (page load speeds, resource
usage, handling many users), usability (intuitive navigation, responsive controls, design), and portability (ease
of deployment on other servers or in other locales) were the main dimensions that were examined in order to
assess its quality using the ISO 25010 standard. Both "technical" and "user" respondent groups evaluated the
system: technical respondents emphasized functionality, dependability, and performance, particularly the
accuracy and pertinence of recommendations made possible by TF-IDF and analytics, while users prioritized
usability, acceptability, and enhancements to day-to-day operations (particularly streamlining product search and
enhancing customer satisfaction). The evaluation provides a fair perspective because the respondents are a
combination of about 60% users and 40% technical stakeholders. Although there is room for improvement,
overall, the results show that RideSmart is technically sound, practically helpful, and effective in real-world
situations. Among the suggested enhancements are enhancing and preserving theunderlying data (for both
products and users), integrating diagnostic or external tools for more precise compatibility, allowing
recommendation adjustments based on a customer’s mechanical expertise, broadening the product database,
adding multilingual support, incorporating user feedback into the recommendation loop, and creating a
mobile‑friendly version to extend reach and usability.
Keywords: Recommendation system, PHP, MySQL, ISO 25010, Agile SDLC, RideSmart, Web-Based,
Descriptive Analytics, System Efficiency.
INTRODUCTION
The motorcycle industry has experienced robust growth, fueled by riders’ enthusiasm for customizing their bikes
to elevate performance, comfort, and style. A booming aftermarket parts market — covering upgrades like
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 911
exhausts, suspension kits, engine modifications, brakes, and aerodynamic components — gives riders vast
options. Yet the overwhelming variety often makes it difficult for individuals, especially those with limited
technical knowledge, to determine which products best suit their specific models and riding goals. Traditional
selection methods—manual research, trial-and-error, or generic advice—can be time-consuming, inefficient, and
even risky, underscoring the need for a better-matched recommendation system.
Enter RideSmart, an intelligent recommendation platform crafted for JavidsonMotorshop’s clientele. It marries
natural language processing, specifically TF IDF, with descriptive analytics to align customer-inputted
preferences (like desired ride feel or performance attributes) with detailed product data. By parsing and weighing
relevant keywords from user text, RideSmart measures compatibility against product specs, while also
accounting for trends gleaned from historical sales and browsing behavior. Built on a content based filtering
model focused on item attributes rather than behavioral data, the system offers personalized, technically sound
parts suggestions. This not only enhances the customization experience—helping riders safely make data-driven
decisions—but also supports business intelligence efforts at JavidsonMotorshop, improving efficiency, customer
satisfaction, and competitive positioning.
To address this gap, RideSmart proposes an intelligent solution that leverages machine learning techniques to
guide riders in selecting the right products for their needs. The system employs a content-based filtering
approach, which analyzes the detailed attributes of both motorcycle models and performance parts such as
technical specifications, compatibility, and performance metrics to generate personalized product
recommendations. By focusing on the built-in features of each item, the system avoids reliance on large volumes
of user interaction data and instead provides tailored suggestions based on the user’s motorcycle and their
specific performance objectives.
This research aims to design, develop, and evaluate the effectiveness of the Ride Revolution system in improving
the decision-making process for riders. The study explores key aspects such as feature extraction, similarity
computation, and the integration of domain knowledge in the recommendation algorithm. Ultimately, RideSmart
seeks to enhance the motorcycle customization experience, empowering riders to make informed, data-driven
choices that lead to safer and more effective performance upgrades.
In one study, Johari, M. Z. F., &Laksito, A. D. (2021). The Hybrid Recommender System of the Indonesian
Online Market Products using IMDb weight rating and TF IDF. This study describes a hybrid recommendation
engine combining TF IDF features (from product descriptions) with IMDb-style weight ratings, using both
demographic filtering and content based analysis to produce more relevant suggestions in an Indonesian e
commerce setting. Another study is [Anonymous authors] (2021). Decision Making System in Online
Marketplace using TF IDF Algorithm in Indonesia: A Micro Analysis of Vespa Spare Parts. This study focuses
specifically on recommending spare parts for Vespa (motor scooter) via content-based filtering powered by TF
IDF. The system matches buyer queries to product descriptions to improve relevance and ease of search in a
niche marketplace setting
Cognitive abilities refer to an individual’s overall mental capacity—encompassing skills like reasoning,
planning, and problem solving. They enable one to engage in abstract thinking, grasp intricate concepts, pick up
new information swiftly, and learn from past experience.
Founded with a mission to deliver quality motorcycles and exceptional service, JavidsonMotorshop has grown
steadily over the years. However, the company now faces modern challenges, including increasing competition,
a rapidly expanding product catalog, and evolving customer expectations in the digital age. The manual approach
to customer assistance has proven inefficient for keeping up with the growing volume of product inquiries and
personalized needs. These limitations highlight the urgent need for an intelligent recommendation system like
RideSmart. By implementing this solution, JavidsonMotorshop aims to enhance operational efficiency, improve
customer satisfaction, and stay competitive in a tech-driven marketplace.
By relying on a fully manual system to handle customer inquiries, JavidsonMotorshop has struggled to manage
the increasing volume of questions about products and individual customer needs. This strain underscores the
need for a smarter solution—like RideSmart—that leverages intelligent recommendations. Adopting this
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 912
technology will help JavidsonMotorshop streamline day-to-day operations, boost customer satisfaction, and
maintain a competitive edge in today’s technology-driven market.The primary objective of this study is to
develop and implement a content-based recommendation system for Ride Revolution, an online motorcycle
shop, to enhance customer experience and improve motorcycle performance customization.
Specific Objectives
• To design and develop a recommendation algorithm that suggests motorcycle parts, accessories, and
upgrades based on user preferences, purchase history, and product features.
• To improve the accuracy and relevance of product suggestions by leveraging content-based filtering
techniques, focusing on item attributes such as compatibility, performance benefits, and customer
reviews.
• To increase user engagement and satisfaction by providing personalized shopping experiences tailored
to individual rider needs and motorcycle types.
• To optimize motorcycle performance by guiding users toward appropriate parts and enhancements that
align with their specific riding style and vehicle specifications.
• To use ISO 25010 standards in the evaluation of the web-based system in terms of:
1. Functional Suitability
2. Usability
3. Reliability
4. Security
5. Maintainability
6. Portability
Scope
The following are the scope of the study:
• Integration of a content-based filtering algorithm into the Ride Revolution e-commerce platform.
• Collection and utilization of product data (brand, type, compatibility, specifications) and user input
(browsing history, previous purchases, preferences).
• Personalized product recommendations aimed at improving both user experience and motorcycle
performance.
• Evaluation of the system’s performance using key metrics such as recommendation accuracy, relevance,
and user satisfaction.
• Recommendation of products based on the algorithm and specifications of the motorcycle.
• Personalization form where the user can fill up the brand, and other specifics of their motorcycle.
• User registration and login (Users can register using Email and password)
• Shopping cart and checkout process (Add products to cart, update product quantity, and remove
individual products to cart)
• Order tracking and history (a list of all past orders with order date and total amount.)
List of Reports Generated within the System:
• products – count of all items in the catalog
• motorcycle_types – count of motorcycle type records
• users – total registered users
• sales – amounts & status used for the Sales card, revenue bar-chart, and order-status donut
LIMITATION
The limitations are:
• The accuracy of recommendations is highly dependent on the completeness and consistency of the
product and user data.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 913
• The system recommends parts based on general compatibility and performance indicators but does not
account for real-time technical diagnostics or user mechanical expertise.
THEORETICAL FRAMEWORK
The RideSmart recommendation system for JavidsonMotorshop uses TF-IDF to extract key features and
keywords from customer reviews and searches, enabling personalized motorcycle product suggestions that
improve purchase decisions and customer experience (Huang, 2025). Descriptive analytics complements this by
analyzing historical sales, browsing behavior, and product interaction data to identify trends and patterns,
supporting data-driven marketing and inventory management (Harvard Business School Online, 2021).
Together, these approaches allow the system to tailor recommendations to individual preferences, enhancing
sales efficiency and operational effectiveness (Sigma Computing, 2025; MEEGLE, 2025).
Figure 1: The Theoretical Framework
The diagram presents a personalized motorcycle product recommendation system that uses TF-IDF and
descriptive analytics to generate tailored suggestions. It starts by collecting customer data such as reviews,
searches, and sales records. The TF-IDF analysis identifies key product features, while descriptive analytics
interprets historical data to reveal trends. These combined insights feed into the recommendation engine, which
outputs personalized product recommendations to enhance sales and customer satisfaction.
CONCEPTUAL FRAMEWORK
Figure 2: The Conceptual Framework
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 914
This diagram represents a conceptual framework for a personalized motorcycle product recommendation system.
Let’s break it down step by step:
User Profile: Contains information about the user, such as motorcycle specifications (bike model, engine type,
etc.) and performance goals (speed, fuel efficiency, style preferences). This acts as the input that guides the
recommendation process.
Feature Matcher (Content-Based Filtering Engine): This module analyzes the user profile and matches it
with the features of available products. Uses content-based filtering, which means it compares attributes (like
specifications or tags) of products with the user’s preferences to find relevant matches.
Product Database: Stores all motorcycle-related products with detailed features, specifications, and
compatibility tags. Serves as the source of data for recommendations.
Recommendation Engine: Receives input from the feature matcher and the product database. Ranks and
suggests products that best match the user’s profile. Essentially, it decides which items are most suitable for the
user.
User Interface: Displays the recommended products to the user.
Allows the user to give feedback, which can further improve future recommendations.
Significance of the Study
This study is significant as it evaluates the effectiveness, usability, and overall quality of the Personalized
Motorcycle Product Recommendation System, providing insights for both practical application and future
improvements.
Use Categories: Users: Shop owners, staff members, and system users (motorists and IT/CS students) benefit
from the study by understanding how the system meets their daily operational and purchasing needs.
Technical People: IT professionals gain insights into the system’s technical performance, reliability, and
portability, helping guide maintenance and optimization strategies.
Future Research: The study serves as a reference for future researchers to enhance recommendation systems,
explore new algorithms, and address emerging user and technical requirements.
Review of Related Literature
Globally, the motorcycle industry continues to grow, driven by increasing urbanization, rising fuel prices, and
the demand for efficient transport in both developed and developing countries (Motorcycle Data, 2023). In
parallel, the market for performance-enhancing motorcycle products has expanded, as riders seek improvements
in speed, handling, fuel efficiency, and aesthetics. The complexity and variety of available products have
increased the need for intelligent systems that can assist riders in making informed choices (Bianchini et al.,
2020).
Several international e-commerce platforms, such as RevZilla in the United States and Motoin in Europe, already
use basic recommendation engines to suggest motorcycle gear and performance parts. However, these systems
often lack deep personalization, particularly when it comes to nuanced rider preferences or specific performance
goals (Chen & Zhang, 2020). As such, a more sophisticated content-based recommendation system could fill
this gap by integrating user behavior, product reviews, and detailed technical specifications.
To overcome the limitations of individual recommendation approaches, hybrid systems combine multiple
techniques, such as content-based and collaborative filtering. These systems leverage the strengths of each
method to provide more accurate and diverse recommendations. For example, a hybrid recommendation system
integrating content-based filtering with collaborative prediction using artificial neural networks demonstrated
improved accuracy and precision in product recommendations (Shao, 2022). Moreover, hybrid systems have
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 915
been effective in addressing challenges like data sparsity and cold-start problems by utilizing both user behavior
data and item attributes (Çano&Morisio, 2019).
The Philippine motorcycle industry has experienced significant growth in recent years, with motorcycles
becoming a preferred mode of transportation due to their affordability and efficiency in navigating congested
urban roads (Land Transportation Office [LTO], 2023). According to the Department of Trade and Industry
(DTI), the motorcycle market in the Philippines has seen an average annual growth rate of 7%, driven by both
personal and commercial usage (DTI, 2022). This increase has led to a rising demand for performance-enhancing
motorcycle parts and accessories.
Content-based filtering has shown potential in improving user satisfaction in recommendation systems by
leveraging product metadata and user preferences (De Guzman & Reyes, 2022). In the context of motorcycle
product recommendations, this means analyzing features such as brand, model compatibility, performance
specifications, and user reviews. A locally-developed system that incorporates this method can offer targeted
recommendations, especially when integrated with indigenous knowledge from Filipino riders' behavior and
preferences.
The integration of recommendation systems in digital platforms enhances user experience by providing
personalized suggestions. Lecaros and Khan (2022) developed Project ATHENA, a hybrid recommendation
engine combining content-based filtering and collaborative filtering techniques. Although their study focused on
academic resources, the methodologies employed can be adapted for recommending motorcycle products based
on user preferences and behaviors.
Synthesis
The motorcycle industry, both globally and locally, continues to grow due to urbanization, rising fuel costs, and
the increasing demand for efficient transport. Alongside this growth is the rising interest in performance-
enhancing motorcycle parts, creating a need for intelligent recommendation systems to help users navigate a
complex and expanding product market. While international platforms like RevZilla and Motoin offer basic
recommendation engines, they often lack deep personalization, prompting the development of more advanced
content-based and hybrid systems. Studies have shown that combining content-based filtering with collaborative
techniques—especially using AI models like neural networks—can enhance recommendation accuracy, solve
data sparsity issues, and adapt to user behavior over time.
Research from countries such as Malaysia and the United States demonstrates the effectiveness of content-based
systems in automotive and motorcycle-related recommendations. These systems prioritize compatibility,
technical specifications, and evolving user preferences—factors highly relevant to personalized motorcycle
recommendations. Locally, the Philippine motorcycle market has seen significant expansion, increasing the
demand for smarter product recommendation tools. Content-based filtering has proven valuable in improving
user experience, especially when tailored to Filipino riders’ behavior and preferences. Projects like ATHENA
and Viajefy highlight the adaptability of hybrid and TF-IDF–based systems in local contexts, while descriptive
studies on consumer satisfaction further support the integration of product quality, pricing, and promotion in
predictive modeling. Overall, both foreign and local findings reinforce the potential and relevance of
RideSmart’s approach, which blends TF-IDF, user behavior, and descriptive analytics to offer personalized
motorcycle product recommendations.
METHODOLOGY OF THE STUDY
Research Design
This study employs both developmental and applied research designs to create and evaluate the RideSmart
recommendation system. The developmental phase focuses on designing, implementing, and testing a prototype
that applies content-based filtering to generate accurate and personalized motorcycle product recommendations.
Guided by the ISO/IEC 25010 software quality model, the system is assessed based on functionality, usability,
reliability, and performance efficiency to ensure it meets user needs and quality standards.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 916
Data Collection
Primary data were gathered directly from Javidson Motorshop’s management, staff, and customers through
interviews, surveys, and observations. Interviews provided insights into the sales process, customer interactions,
and common challenges in offering personalized product suggestions. Surveys and observations helped identify
customer preferences, buying behavior, and browsing patterns, which served as the foundation for designing the
recommendation system. After the system development, the personalized motorcycle product recommendation
system was evaluated based on ISO 25010 standards. Its functionality, reliability, efficiency, usability, and
portability were assessed to ensure it met user and business requirements effectively. The evaluation confirmed
that the system could provide accurate recommendations, operate consistently, perform efficiently, and offer a
user-friendly and portable experience.
Software Methodology
Figure 3: SDLC Agile Model
The development of the system follows the System Development Life Cycle (SDLC) using the Agile
methodology to ensure a structured yet flexible approach. Each phase—planning, analysis, design,
implementation, testing, and maintenance—plays a crucial role in building a robust and efficient system. The
planning and analysis stages focus on gathering requirements and understanding user needs to guide effective
system design. During the design and implementation phases, developers construct the system architecture, write
code, and integrate features to achieve project goals. Finally, through continuous testing and maintenance, the
team ensures the system remains stable, reliable, and adaptable for long-term use.
Figure 4: Use Case Diagram
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 917
UML Diagram: diagram illustrates the system’s structural and behavioral design, showing how different
components interact within the personalized recommendation system. It includes use case, class, and sequence
diagrams that define user interactions, data flow, and system processes. This visual representation ensures clear
communication between developers and stakeholders, promoting organized and efficient system development.
Figure 5: Data Flow Diagram
Data Flow Diagram (DFD): presents the logical flow of information within the recommendation system, from
user input to product recommendation output. It highlights how data is processed through components such as
the user profile, feature matcher, and recommendation engine. The DFD ensures a clear understanding of data
movement, system boundaries, and functional relationships among processes.
Database Function: The database serves as the central repository that stores user profiles, motorcycle
specifications, and detailed product information for analysis and recommendation. It supports data retrieval for
TF-IDF processing and descriptive analytics, ensuring accurate and efficient generation of personalized
suggestions. The database design prioritizes data integrity, scalability, and quick access to enhance the overall
system performance and reliability.
Respondents
A total of 101 respondents participated in the evaluation of the system based on ISO 25010 standards. The group
consisted of 1 shop owner, 2 staff members, 88 users (motorists and IT/CS students), and 10 IT technical
professionals. They assessed the system’s functionality, reliability, efficiency, usability, and portability to ensure
it met both user expectations and technical standards.
Development and Evaluation Procedure
Development Tools
Throughout the system's development, the following technologies and tools were used for version control,
coding, testing, and deployment, which together defined the development environment:
• Development Environment Frameworks: PHP, HTML, CSS
• Database: MySQL Full-Stack
• Framework: Livewire Code
• Editor: Visual Studio Code
• Survey: Google Form
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 918
Hardware Used / System Requirements The hardware listed below supported the development of the
Personalized Motorcycle Product Recommendation System:
• Minimum Computer Specifications
• Processor: Intel or AMD, minimum 2.0 GHz Dual-Core
• RAM: 4 GB or higher
• Storage: 500 GB HDD or SSD
• Display: 720p resolution (1280 x 720 pixels) or higher
• Connectivity: Web browser with internet access
Evaluation Process
Evaluation is the process of completely appraising a product by acquiring and analyzing information about its
actions, characteristics, and results. Forming opinions about the product, improving its efficacy, and assisting in
related decision-making are the objectives. Patterson (1987). Two methods of evaluation were used in the study.
While the second looked at the application's overall usefulness and technical performance, the first concentrated
on gauging user happiness and simplicity of use. Despite using two approaches, the assessment is conducted
using a single evaluation form that is based on ISO 25010 standards. The system is assessed in this study to meet
the ISO 25010 requirements, including:
Functionality
• Accuracy of Functionality: Are all product details, prices, and inventory accurate?
• Completeness of Functions: Can users browse, search, add to cart, and complete checkout?
Reliability
• Availability: Is the website accessible 24/7 without downtime?
• Fault Tolerance: Does the system recover gracefully from unexpected errors?
Efficiency
• Response Time: Does each page load within an acceptable time?
• Resource Utilization: Is the website optimized to use minimal server and bandwidth resources?
• Scalability: Can the system handle a large number of concurrent users?
Usability
• Learnability: Can a first-time user navigate the site easily?
• Operability: Are interactive elements like buttons and forms intuitive and responsive?
• User Interface Aesthetics: Is the design clean, appealing, and brand-consistent?
Portability
• Installability: Can the website be deployed easily on another server or host?
• Adaptability: Can the system be customized for new locations or services?
Data Analysis Plan
The statistical instruments employed in the study to guarantee precise and insightful analysis of the gathered
data are described in this part. The study employed the following statistical tools: 1. The weighted mean A
statistical method for gauging the degree of agreement or disagreement among different questionnaire questions
or assertions is the weighted mean, sometimes referred to as the average mean. Each item's weighted mean is
determined by multiplying its weighted points by the matching frequency or sample size, then adding the results.
Next, this sum is split by the total number of responders.
In quantitative research and evaluations, a Likert scale is a data gathering technique used to gauge people's
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 919
attitudes, beliefs, or views. In order to facilitate statistical analysis, it asks respondents a sequence of statements
or questions, each of which has a set of response possibilities with numerical values assigned to them. A 4-point
Likert scale is used in this study, with values allocated from 4 for "strongly agree" to 1 for "strongly disagree."
Normally, a Likert scale goes from "strongly agree" to "strongly disagree." The table below shows the 4-point
Likert scale used in the study.
The System
The study developed an application called RideSmart: A Personalized Motorcycle Product Recommendation
System Using TF-IDF and Descriptive Analytics for JavidsonMotorshop, designed to provide customers with
personalized motorcycle product recommendations. By utilizing TF-IDF and descriptive analytics, RideSmart
analyzes customer preferences, purchase history, and product information to suggest the most suitable parts,
accessories, or services, offering a more efficient and satisfying shopping experience. It also assists
JavidsonMotorshop in understanding customer behavior by identifying buying trends, popular products, and
demand patterns, helping improve inventory and marketing strategies. Built using Laravel (PHP Framework),
Python, MySQL, HTML, and CSS, the system was assessed using ISO 25010 standards, ensuring quality in
functionality, reliability, efficiency, usability, and portability.
1. User Dashboard (Revenue Overview) – This interface displays the motorshop’s revenue from the past
few months, giving administrators a clear and organized view of financial performance. It helps in
tracking sales trends and making data-driven business decisions.
Figure 1.
2. Product Management Panel – This section shows all available motorcycle products, allowing staff to
easily monitor inventory, update product details, and manage stock levels efficiently to ensure smooth
operations and customer satisfaction.
Figure 2.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 920
Assessment: Summary Of Respondents On The System.
Table 1.
The ISO 25010 evaluation table demonstrates that the Personalized Motorcycle Product Recommendation
System was highly rated by 101 respondents, including both users and technical experts. All evaluated criteria—
Functionality, Reliability, Efficiency, Usability, and Portability—received “Strongly Agree” ratings, indicating
that the system performs its intended functions correctly and consistently. Users and experts both found the
system efficient, completing tasks within acceptable time and resource constraints, while also being user-friendly
and easy to navigate. Portability received slightly higher ratings from technical experts, suggesting confidence
in the system’s adaptability to different environments. Overall, the system’s average ratings confirm it meets
ISO 25010 quality standards and is considered highly acceptable by both user groups.
Ethical Considerations
The study guarantees the confidentiality and integrity of the data obtained from participants. Information
provided by respondents is secure, and no personally identifiable information is shared without permission.
Respondents are allowed to leave the study at any moment without facing any repercussions, in accordance with
the principles of voluntary participation. Strict adherence to data security protocols guards against misuse and
illegal access to information. Lastly, in order to preserve the study's integrity, all results are presented truthfully
and accurately, free from prejudice or manipulation.
Summary
To provide more precise motorcycle part recommendations, a web-based application called the Personalized
Motorcycle Product Recommendation System was created. The primary aim of this system is to precisely suggest
to motorbike owners which parts are suitable with their vehicles. Users and technical respondents assess the
application's acceptability and usefulness using the ISO 25010 evaluation format. Because this service develops
an online recommendation system with an intuitive design, it is important for both motorbike users and
motorshop owners. For the client, it lowers the chance of motorbike parts being incompatible.
CONCLUSION
In summary, the comparison of the technical and user respondents' assessments demonstrates that the system is
thoroughly evaluated, placing equal weight on its technical performance and practical applicability. The user
group highlighted the system's acceptance, usefulness, and benefits for day-to-day shop operations, pointing out
Criteria
(ISO25010)
Respondents (101 )
Users (91) staff & owner included
Technical (10 )
WM
VI
WM
VI
1. Functionality
3.53
Strongly Agree
3.53
Strongly Agree
2. Reliability
3.54
Strongly Agree
3.58
Strongly Agree
3. Efficiency
3.54
Strongly Agree
3.54
Strongly Agree
4. Usability
3.55
Strongly Agree
3.48
Strongly Agree
5. Portability
3.43
Strongly Agree
3.68
Strongly Agree
Overall Average Mean
3.52
Strongly Agree
3.56
Strongly Agree
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 921
how it improves customer experiences and streamlines product searches. The technical respondents, however,
concentrated on assessing the system's overall effectiveness, functionality, and dependability. They verified that
the system's TF-IDF and analytics functions provide precise and pertinent recommendations. The results offer a
well-rounded perspective because of the balanced representation, which includes 60% user respondents and 40%
technical respondents. Overall, the results demonstrate that the system is not only sound technically but also
advantageous and successful in practical applications, making it a useful tool for JavidsonMotorshop and other
shops of a similar nature. They also highlight areas that could use more development and improvement.
RECOMMENDATION
To enhance RideSmart's capabilities, integrating hybrid recommendation models combining collaborative
filtering with machine learning algorithms would significantly improve prediction accuracy beyond TF-IDF
limitations. Incorporating real-time diagnostic data through VIN-based compatibility checking and continuous
user feedback loops would strengthen personalization and system intelligence. Expanding implementation across
multiple motorshops and diverse motorcycle brands would validate generalizability and robustness across varied
market contexts. Developing mobile-responsive interfaces is essential for contemporary user accessibility and
on-the-go purchasing scenarios. Finally, conducting rigorous comparative analyses against established
recommendation systems using standardized metrics (precision, recall, conversion rates) would provide
empirical validation of RideSmart's competitive advantage and justify broader industry adoption.
By improving the accuracy of its recommendations through more thorough and consistent product and user data
management—such as incorporating frequent data updates and validation procedures—"RideSmart: A
Personalized Motorcycle Product Recommendation System" should be improved in order to overcome its present
limitations. The system should be improved to incorporate real-time technical diagnostics or integration with
external diagnostic tools for more accurate and situation-specific recommendations in order to get beyond the
dependence on universal compatibility. It may also be possible to incorporate features that take into consideration
the user's degree of mechanical competence, providing both basic and advanced recommendations suited to
various user profiles. Expanding the product database, adding language support for greater accessibility,
integrating user comments into the suggestion process, and creating a mobile-friendly version of the system to
increase usability and reach are additional potential enhancements. These improvements would optimize the
system's total value for both technical and non-technical applications, in addition to improving its dependability.
REFERENCES
1. Foreign Studies Rahim, M. S., Shiblee, M. N. I., & Asaduzzaman, M. (2020). A recommendation system
for automotive spare parts using content-based filtering. International Journal of Advanced Computer
Science and Applications, 11(1), 621–627. https://doi.org/10.14569/IJACSA.2020.0110177
2. Suryanarayana, A., Bhuvaneswari, R., & Dinesh, B. (2021). Spare part recommendation system for
vehicle repair using content-based filtering. Journal of Ambient Intelligence and Humanized Computing,
12(10), 9285–9295. https://doi.org/10.1007/s12652-020-02463-
3. Pazzani, M. J., & Billsus, D. (2007). Content-based recommendation systems. In P. Brusilovsky,
A. Kobsa, & W. Nejdl (Eds.), The adaptive web (pp. 325–341). Springer. https://doi.org/10.1007/978-3-
540-72079-9_10
4. Local Studies Arruejo, M. C., & Arruejo, D. C. (2025). Development and evaluation of Viajefy: A
tourism information system using TF-IDF algorithm. Mindanao Journal of Science and Technology,
23(1). Retrieved from https://doi.org/10.61310/mjst.v23i1.2383
5. Babia, C. J., Garcia, M. R., & Taparan, J. F. (2024). Analyzing the customer satisfaction of
F. Taghap Motorcycle Parts Shop in General Trias, Cavite. Retrieved
from https://www.researchgate.net/publication/378430448_ANALYZING_THE_CUSTOMER_SATIS
FACTION_OF_F_TAGHAP_MOTORCYCLE_PARTS_SHOP_IN_GENERAL_TRIAS_CAVITE
6. Foreign Literature Bianchini, M., Gori, M., &Scarselli, F. (2020). Recommender systems in the
automotive aftermarket: A machine learning approach. International Journal of Data Science and
Analytics, 9(4), 213–225. https://doi.org/10.1007/s41060-020-00219-x
7. Chen, X., & Zhang, H. (2020). User satisfaction in motorcycle e-commerce: A recommender system
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue X October 2025
www.rsisinternational.org
Page 922
perspective. Journal of Retailing and Consumer Services, 54,
102001. https://doi.org/10.1016/j.jretconser.2019.102001
8. Cano, E., & Morisio, M. (2017). Hybrid recommender systems: A systematic literature review.
Intelligent Data Analysis, 21(6), 1487 1524. https://doi.org/10.48550/arXiv.1901.03888
9. Local Literature Land Transportation Office (LTO). (2023). Motor vehicle registration
statistics. https://www.lto.gov.ph
10. Department of Trade and Industry (DTI). (2022). Philippine motorcycle market report
2022. https://www.dti.gov.ph
11. De Guzman, F. R., & Reyes, L. M. (2022). Implementation of a content-based recommendation engine
in Filipino e-commerce. Philippine Journal of Computing, 19(1), 45–56.
12. Lordjette Leigh M. Lecaros, Concepcion L. Khan (2022). A Tech Hybrid-Recommendation Engine and
Personalized Notification: An integrated tool to assist users through Recommendations (Project
ATHENA) https://doi.org/10.48550/arXiv.2202.06248
13. Huang, S. (2025). Personalized intelligent recommendation algorithm for consumer purchase behavior.
International Journal of Information Technologies and Systems Approach. Retrieved from
https://www.researchgate.net/publication/393881583_Personalized_Intelligent_Recommendation_Algo
rithm_for_Consumer_Purchase_Behavior
14. Harvard Business School Online. (2021). What is descriptive analytics? 5 examples. HBS Online.
Retrieved from https://online.hbs.edu/blog/post/descriptive-analytics
15. Sigma Computing. (2025). If you're not using TF-IDF in data analysis, you're missing out. Sigma
Computing. Retrieved from https://www.sigmacomputing.com/blog/tf-idf-definition
16. MEEGLE. (2025). Recommendation systems for descriptive analysis. MEEGLE. Retrieved from
https://www.meegle.com/en_us/topics/recommendation-algorithms/recommendation-systems-for-
descriptive-analysis
