STUDENT PLACEMENT ANALYZER:

A RECOMMENDATION SYSTEM USING MACHINE LEARNING

Literature Review

Author : APOORVA RAO R , DEEKSHA K C, VISHAL PRAJWAL, VRUSHAK K, NANDINI M S

 This literature review focuses on the application of machine learning algorithms, specifically Fuzzy C
Means Clustering and Naive Bayes Algorithm, in predicting student placements in educational
institutions.

 The purpose is to critically evaluate the existing literature on these algorithms, their effectiveness, and
their limitations in the context of student placement prediction.

 The review is structured into key sections: Background, Theoretical Framework, Research Gap
Identification, Methodology, Main Themes, Synthesis and Analysis, Discussion, and Conclusion.

1. Background and Context:

 Student Placement Prediction is an automated system that leverages machine learning by evaluating
various attributes such as knowledge, skills, and attitudes of the student. Fuzzy C Means Clustering is
a specific algorithm used to group students based on these attributes.

 The need for automated student placement systems has grown with the increasing complexity and
competitiveness of the job market. Traditional methods are often manual and less efficient.

2. Theoretical Framework:

 The theoretical framework for this topic is grounded in machine learning algorithms, specifically
Fuzzy C Means Clustering and the Naïve Bayes Algorithm. These algorithms are employed in a
sequential manner to first cluster students based on their attributes and then classify them into
different placement statuses.

 Fuzzy C Means provides multiple clusters, thereby capturing the complexity of student attributes, On
the other hand, the Naïve Bayes Algorithm offers a probabilistic framework for classification,
making it suitable for predicting placement statuses based on the clusters formed.

3. Research Gap Identification:

 The existing literature primarily focuses on manual methods for predicting student placements, which
are time-consuming, less efficient, and result in lower user satisfaction.

 The research questions this review aims to address include: How can machine learning improve the
efficiency and accuracy of student placement predictions? What are the limitations of current manual
methods, and how does the proposed system overcome these limitations?
 4. Methodology:

 I began my search on the IJARIIE journal's website, focusing on machine learning applications in
student placement recommendation systems.

 The paper by Apoorva Rao R and team was chosen because it offers a unique approach to student
placements by categorizing them into five different statuses. This approach aligns perfectly with my
research focus on the complexities of student placements and the role of machine learning in
addressing them.

5. Main Themes and Trends:

 Key themes include the use of historical data as the training dataset, the application of Fuzzy C Means
for clustering, and the Naïve Bayes Algorithm for classification.

 The main theme revolves around The Fuzzy C Means is employed for clustering students based on
their attributes like knowledge, skills, and attitudes. Following clustering, the Naive Bayes Algorithm
is applied for classification, categorizing students into different placement statuses.

6. Synthesis and Analysis:

 It contrasts traditional manual methods, which are characterized by time consumption and less user
satisfaction, with an automated, data-driven approach.

 The paper employs Fuzzy C Means Clustering and the Naïve Bayes Algorithm as its core
methodologies, allowing for a more complex understanding of student attributes, However, The paper
leverages historical data as the training dataset, a practice that adds robustness but also raises
questions about potential biases or limitations.

 Hence, It connects with broader trends in educational research but leaves room for further
comparative studies and analysis of potential biases in using historical data.

7. Discussion:

 By transitioning from traditional manual methods to an automated system, institutions can achieve
increased efficiency, accuracy, and user satisfaction.

 Additionally, I plan to conduct a comparative analysis with other machine learning techniques such as
Decision Trees and Random Forest to evaluate their efficacy in this context.

 The paper also cites other works in educational research, such as "Study of Implementation of Online
Placement System" and "Performance Prediction of Students Using Distributed Data Mining," but
does not engage in a comparative analysis. This lack of discussion on disagreements or debates within
the literature.
Conclusion:

 In summary, the paper provides valuable insights into the automation of student placement processes
but leaves room for further exploration and validation.

 The significance of this review lies in its exploration of machine learning applications in the
educational sector, a domain traditionally reliant on manual processes senarios.

 Future research could focus on these areas, providing a more comprehensive understanding of the best
practices in student placement prediction. This could include investigating the limitations of using
historical data.

References:

 [1]. Fatin Aliah Phang : Postgraduate Supervision: Supervisors versus Students .

MOHAK BATRA

20841A1210

Guided By :

Ms. Subhashini