https://doi.org/10.46335/IJIES.2024.9.8.

17 e-ISSN: 2456-3463
Vol. 9, No. 8, 2024, PP. 77-80
International Journal of Innovations in Engineering and Science, www.ijies.net

College Recommendation System

Ashish Patle, Gouri Ghushe, Bhushan Meshram, Mandar Bhole, Dr. K. O. Gupta

Department of Computer Engineering’s. Vincent Pallotti College of Engineering & Technology,Nagpur, India

[email protected]

Received on: 05May, 2024 Revised on: 03 July, 2024 Published on: 06 July, 2024

the K-Nearest Neighbors (KNN) algorithm. This approach

Abstract: Selecting a college can be daunting, especially with aims to offer personalized recommendations based on factors
limited information about individual colleges. Despite strong such as faculty quality and campus life. KNN plays a crucial
academic performance, students may choose colleges that role in identifying colleges similar to user preferences,
don't fully align with their needs due to a lack of streamlining the decision-making process for students.
comprehensive data. This paper proposes a novel college Overall, integrating KNN into recommendation systems
recommendation system to empower students and parents in
making informed decisions.
I.INTRODUCTION: enhances their effectiveness in providing tailored college
options.
Choosing the right college can be really tough with so many
options out there. Luckily, technology has stepped in to help In this Paper [2], This paper discusses the challenges faced by
with something called a college recommendation system. This students after completing their SSC examinations, especially
system uses fancy computer stuff like data analysis and when it comes to choosing the right stream and college. The
machine learning to suggest colleges that match what students focus is on simplifying the junior college admission process
are looking for. It considers things like college fees, ratings, and assisting students in finding colleges that align with their
how good the teachers are, the social scene, job opportunities needs. To achieve this, the authors utilize data mining and
after graduation, and what subjects students like. By using a query optimization techniques to recommend suitable
special algorithm called K-Nearest Neighbors, it finds colleges. Additionally, they employ algorithms like Naive
colleges that are similar to what students want, and it Bayesian and Decision trees to improve recommendation
organizes them neatly, often by state, to make them easy to accuracy and minimize search time, ultimately providing
look through. Not only that, but it also helps students keep students with optimized college choices. By integrating these
track of upcoming exams and deadlines so they can stay on methods, the paper aims to alleviate the stress and uncertainty
top of things. associated with college selection, empowering students to
make informed decisions about their educational futures.
By showing colleges based on where they are, it makes it
easier for students to find options that are close by. This paper In this Paper [3], authors state that in the current era, students
explores how these recommendation systems work, how face challenges in selecting colleges that offer excellent
effective they are, and how they could change the way we educational standards, campus facilities, placements, and
choose colleges. We'll look at both the good and the not-so- other desirable features. Lack of proper information often
good sides of these systems to understand their impact on leads to students missing out on admission to their desired
higher education better. colleges or branches. To address this issue, a recommendation
system is proposed based on factors like college accreditation
II.LITERATURE SURVEY: grades, campus placements, and alumni feedback. By
incorporating Semantic analysis algorithms to capture
In this Paper [1], Prof. Sunil Sonawane, Miss. Madhavi sentiments and combining Naive Bayes and Adaboost
Gosavi, Miss. Aparna Pawar, Mr. Omkar Pawar, Mr. algorithms for ranking, the system aims to assist students and
Dhananjay Dighe, In exploring college recommendation parents in making informed choices. The architecture includes
systems, . proposed utilizing AI and ML techniques, including modules for admin, alumni, and user functionalities, ensuring

77
https://doi.org/10.46335/IJIES.2024.9.8.17 e-ISSN: 2456-3463
Vol. 9, No. 8, 2024, PP. 77-80
International Journal of Innovations in Engineering and Science, www.ijies.net

seamless operation. Overall, this recommendation system IV.PROPOSED METHODOLOGY:

streamlines the college selection process, making it easier for
users to identify suitable colleges based on their preferences Our upcoming website is like a helpful guide for students
and requirements. looking to join engineering colleges. We're building it using
simple web languages like HTML, CSS, and JavaScript,
[4] The project endeavors to rectify the dearth of tailored along with PHP. Imagine it as a digital assistant that uses a
platforms catering to SSC graduates in their pursuit of college special method called KNN (which is a smart way to find
admissions. Through the deployment of a desktop program, similarities) to suggest colleges based on what students want.
we present an intuitive interface equipped with filters to
generate customized lists of universities aligning with To make sure our recommendations are spot-on, we've
students' preferences. By harnessing advanced algorithms and collected data from different sources, like Kaggle. We tidy up
soliciting user feedback, the system furnishes pre-populated this data, making sure it's clean and organized, and then store
college selections predicated on criteria such as caste and it in our system.
geographical location. Comprising four distinct modules, the When a student visits our website, they can tell us what they're
platform facilitates comprehensive college exploration, looking for - things like college fees, faculty rating, and
comparison, and branch-specific inquiries. Future endeavors campus vibe. Our system then matches this info with the
include extending the system's functionality to encompass colleges in our database and shows a list of options, grouped
additional educational streams. Ultimately, our initiative by state. It's like having a personalized advisor right at your
aspires to streamline the college selection process, fingertips.
empowering students to make well-informed decisions.
[5] Disease risk prediction is a significant challenge in the
medical field, often tackled using machine learning Start
algorithms. Among these, the k-nearest neighbor (KNN)
algorithm stands out. This paper explores various KNN
variants and compares their performance for disease
prediction, using eight benchmark datasets from reputable
Construct
sources. The study evaluates accuracy, precision, and recall, comprehensiv
identifying the Hassanaat KNN and ensemble approach KNN e
as top performers. Additionally, the paper delves into the Profile
algorithmic aspects of KNN variants, including adaptive
approaches, weight attribution, and distance metrics. Notably,
it highlights the need for comparative performance analysis
across KNN variants, which this study addresses
comprehensively. Overall, the findings provide valuable Include
insights for healthcare researchers seeking to leverage Include Include
Personal
machine learning for predictive disease analytics. Fees other
Preferenc
pertine
e

III.PROBLEM STATEMENT:

Finding the right engineering college can be challenging for

students. To simplify this process, the College Analyzed Analyzed
Analyzed
Recommendation System analyzes students' grades and Personal Additiona
Integers
preferences to generate a tailored list of suitable colleges. By fit l
automating the search and providing detailed information, it
saves students valuable time and helps them make well-
informed decisions about their education. With this system,
students can confidently choose a college that meets their
academic and personal needs, setting them on the path to
success in their careers, the system saves time and enables Generating
students to make well-informed choices. With this support, College
students can confidently embark on their engineering journey, Recommendatio
equipped with the knowledge to select a college that best n
matches their needs and aspirations. list

78
https://doi.org/10.46335/IJIES.2024.9.8.17 e-ISSN: 2456-3463
Vol. 9, No. 8, 2024, PP. 77-80
International Journal of Innovations in Engineering and Science, www.ijies.net

A. Process: through the dataset, calculating the distance between the

user's input and every college's attributes.
During the initial stage, we'll focus on building both the
visible part (the frontend) and the hidden part (the backend) Once the distances are calculated, the algorithm sorts the
of our system. Once this phase is complete, we'll dive into colleges based on their proximity to the user's input, with the
working on the dataset. This dataset will hold detailed closest colleges ranked higher in the list. The "K" nearest
information about various engineering colleges, including colleges are then selected based on the value of K,
their fees, locations (city and state), available streams, faculty representing the number of nearest neighbors to consider. In
ratings, college ratings, social life ratings, placement ratings, the provided code, K is set to 10 by default.
and more.
Finally, the script presents the user with a list of colleges that
User registration details, login credentials, and all other are closest to their preferences, organized by state and stream.
important information needed by the system will be securely This output allows users to easily identify colleges that
stored on our local server. We'll then thoroughly train and test closely match their criteria, aiding them in making informed
this dataset by putting it through different scenarios. Based on decisions about their education and college selection process.
the results we observe, we'll make any necessary adjustments By leveraging the KNN algorithm and Euclidean distance
or improvements to ensure the system works smoothly and metric, the script efficiently generates personalized
accurately. recommendations for users seeking suitable colleges.
B. Module work: D. Module:

The code in a PHP script designed to handle form In this module, an Application Programming Interface (API)
submissions and generate a list of colleges based on user is employed to access news articles from newsapi.org. The
input. When a form is submitted, the script loads data from a primary objective is to provide students or users with relevant
CSV file containing information about various colleges, such news pertaining to upcoming exams. By integrating this API,
as their names, states, streams, fees, ratings, and other the system aims to enhance awareness among students
attributes. This data is preprocessed to ensure consistency and regarding important events and developments that may
accuracy. impact their academic pursuits. It serves as a valuable
The script then receives input from the form, including the resource for students to stay informed about exam schedules,
user's preferred state, stream, undergraduate fee, and ratings changes in examination patterns, study tips, and other relevant
for factors like faculty, placement, and social life. Using the information. It's noteworthy that the API used in this module
K-Nearest Neighbors (KNN) algorithm, is freely available, although the number of news articles that
the script calculates the similarity between the user's input and can be fetched is limited due to the nature of the free version.
the colleges in the dataset. It identifies the colleges that are Nonetheless, even with this limitation, the system plays a
closest to the user's preferences and organizes them by state crucial role in disseminating timely and pertinent news
and stream. updates to students, thereby contributing to their overall
preparedness and success in their academic endeavors.
Finally, the script displays the list of colleges matching the
user's criteria in a structured format, with each college V.FUTURE SCOPE:
grouped by state and stream. This output provides users with This system offers numerous benefits to students by providing
a concise and organized overview of colleges that align with a comprehensive platform that amalgamates various
their preferences, helping them make informed decisions functionalities. Currently, there is a significant gap in the
about their education availability of a single platform that enables students to search
C. Algorithm: for colleges based on their preferences, both in terms of
specialization and location, while also delivering relevant
The K-Nearest Neighbors (KNN) algorithm utilized in the news updates regarding exams. Although the system currently
provided code is a simple yet effective method for caters to specific college streams such as engineering, law,
classification and regression tasks. At its core, KNN operates science, arts, management, hotel management, etc., our future
by measuring the similarity between data points using a plans include expanding its scope to accommodate additional
distance metric, typically the Euclidean distance. This streams and incorporate more features. This expansion will
distance metric calculates the straight-line distance between ensure that the system becomes even more versatile and
two points in a multi-dimensional space, providing a measure beneficial, catering to the needs of a broader spectrum of
of their similarity or dissimilarity.In the context of the script, students across different educational disciplines. By
the KNN algorithm compares the user's input, representing continuously evolving and enhancing its capabilities, our goal
their preferences for college attributes, with the attributes of is to create a unified platform that simplifies the college
each college in the dataset. This comparison is achieved by selection process and provides valuable resources to students
computing the Euclidean distance between the user's input pursuing various academic paths.
and the attributes of each college. The algorithm iterates

79
https://doi.org/10.46335/IJIES.2024.9.8.17 e-ISSN: 2456-3463
Vol. 9, No. 8, 2024, PP. 77-80
International Journal of Innovations in Engineering and Science, www.ijies.net

VI.CONCLUSION:
The College Recommendation System is designed to alleviate
the burden on students and simplify the college selection
process. By providing a user-friendly interface, the system
offers students a curated list of colleges tailored to their
preferences. To achieve this, we have utilized diverse datasets
and machine learning techniques to ensure accurate and
relevant recommendations. Additionally, the system
incorporates an API to display news related to exams and
upcoming examination schedules, aiming to enhance
awareness among students and their parents. This
comprehensive approach not only facilitates informed
decision-making but also serves as a valuable resource for
learners navigating the complexities of college admissions.
REFERENCE:
[1] Prof. Sunil Sonawane, Miss. Madhavi Gosavi, Miss. Aparna
Pawar, Mr. Omkar Pawar Mr. Dhananjay Dighe” college
recommendation system using artificial intelligence and
machine learning” Volume:05/Issue:03/March-2023
International Research Journal of Modernization in
Engineering Technology and Science e-ISSN: 2582-5208

[2] Vinit Jain, Mohak Gupta, Jenish Kevadia, Prof. Krishnanjali

Shinde, Special Issue – 2017, International Journal of
Engineering Research & Technology (IJERT) ISSN: 2278-
0181 Published by, www.ijert.org ICIATE - 2017 Conference
Proceedings “College Recommendation System”

[3] Miss. Deokate monali, Miss. Gholave Dhanashri, Miss. Jarad

Dipali, Miss. Khomane Tejaswini, Guided By: Prof. Nale R.K,
” college recommendation system for admission”, e-ISSN:
2395-0056 Volume: 05 Issue: 03 | Mar-2018 www.irjet.net p-
ISSN: 2395-0072

[4] Prof. Pankaj Sharma, Sankalp Raj, Safiullah Rayen, Ajeet

Kumar, Abhishek Kumar Bhandari, “Intelligent
Recommendation System for Course Selection in Indian
Higher Education System” International Journal of Advanced
Research in Science, Communication and Technology
(IJARSCT), Volume 3, Issue 14, May 2023, ISSN (Online)
2581-9429

[5] Shahadat Uddin, Ibtisham Haque, Haohui Lu, MohammadAli

Moni & ErgunGide, “Comparative performance analysis of
K‑nearest neighbour (KNN) algorithm and its diferent variants
for disease prediction” , (2022) 12:6256 |
https://doi.org/10.1038/s41598-022-10358-x

80