International Journal of Scientific Research in Engineering and Management (IJSREM)

Volume: 08 Issue: 09 | Sept - 2024 SJIF Rating: 8.448 ISSN: 2582-3930

SMART RESUME ANALYZER

Mr. Ashish Modi1, Mr. Hassan Raza Chowdhary2

1Asst. Prof / Department of Information Technology, Nagindas Khandwala College, Mumbai, Maharashtra, India,
2 Student / Department of Information Technology, Nagindas Khandwala College, Mumbai, Maharashtra, India.

---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In today's competitive job market, the need for streamline the process of matching candidates to job
efficient and effective hiring processes is paramount. The profiles. Some of these platforms have successfully
Smart Resume Analyzer is an innovative AI-powered tool implemented techniques to categorize and rank resumes
reimagining the recruitment landscape by seamlessly based on their relevance to specific job postings. Among
integrating advanced machine learning (ML) and natural the most effective methods are Content-based
language processing (NLP) techniques. Unlike traditional Recommendation systems, Cosine Similarity, and K-
resume screening methods that are often hampered by manual
Nearest Neighbours (KNN) algorithms, which help in
evaluation, subjectivity, and inefficiencies, this system offers a
identifying the resumes that most closely align with a job
cutting-edge approach to candidate assessment. It intelligently
extracts and analyzes key information from resumes, utilizing description.
sophisticated algorithms such as Cosine Similarity and Term
However, despite these advances in accuracy and
Frequency-Inverse Document Frequency (TF-IDF) to provide
real-time, data-driven evaluations. precision, the time required to find suitable candidates
remains a significant drawback. The challenge lies in
Moreover, this tool democratizes the job search process by efficiently and accurately matching each resume to the
empowering applicants with the knowledge and tools to refine appropriate job posting in a way that reduces processing
their resumes, thereby leveling the playing field in a time without compromising the quality of the results. To
competitive job market. The Smart Resume Analyzer is more address this, innovative approaches are needed that not
than just a recruitment aid—it is a transformative solution that
only improve the speed of candidate identification but
bridges the gap between employers and the talent they seek,
also maintain high standards of accuracy and relevance in
fostering a more efficient and equitable hiring process.
matching candidates to job opportunities. These
Key Words: NLP (Natural Language Processing) Resume advancements will be crucial in optimizing the e-
Parsing, Machine Learning, Recommender Systems, Data recruitment process, making it more efficient and
Security, Data Privacy, User Engagement, Semantic Analysis,
User Data Analysis: & Text Mining.
effective for both recruiters and job seekers.

2. REVIEW OF LITERATURE
1. INTRODUCTION
Satyaki Sanyal and his team [1] have developed a resume
Any Recruiters will often face the challenge of analysis software that automates the extraction and
selecting the best candidates from an overwhelming pool evaluation of pertinent information from submitted
of applicants for a given job vacancy. Manually sifting resumes. Their approach integrates Natural Language
through thousands of resumes to identify the most Processing (NLP), Machine Learning (ML), and data
qualified candidates is not only time-consuming but also mining techniques to detect keywords, patterns, and
prone to errors and inconsistencies. Although job trends. By assessing resumes in relation to job
websites have introduced methods to improve accuracy specifications, the software streamlines the recruitment
and precision in candidate matching, a significant process, lightens the load on recruiters, and highlights the
limitation remains in the time it takes to process and most qualified candidates.
compare every resume against multiple job postings. This
Vaidya and colleagues [2] describe a technique used by
high time complexity becomes a bottleneck, especially
resume analyzers to retrieve relevant information. Their
when dealing with vast datasets.
method involves removing stop words and applying the
With over 50,000 e-recruitment platforms emerging in Soundex algorithm to group similar-sounding words.
recent years, various strategies have been developed to

© 2024, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM37514 | Page 1

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 08 Issue: 09 | Sept - 2024 SJIF Rating: 8.448 ISSN: 2582-3930

Identified keywords are then cataloged in a database for job applications. This addresses the challenge of
further processing. manually reviewing a large volume of resumes, which has
become increasingly impractical for recruiters.
Daryania and his team [3] present an automated resume
screening system that leverages NLP to evaluate resumes Thakur and Goyal [10] present a Resume Classification
and rank them based on their alignment with job System (RCS) that uses NLP and ML techniques to
requirements. This system extracts keywords, skills, and automate resume analysis. While their model improves
qualifications from resumes and compares them to job the efficiency and transparency of the screening process,
descriptions, generating a relevance score to assist in it does not provide recommendations for enhancing
ranking. applicant resumes. Their work demonstrates how NLP
and RCS can significantly reduce the recruiter’s
Nawander and associates [4] introduce a method workload and deliver effective results.
combining NLP with Streamlit modules to extract data
from PDF resumes, store it in a database, and analyze it 3. METHODOLOGY
for ranking purposes. Their system also offers
suggestions for enhancing resume presentation, including The development of a smart resume analyzer that focuses
on job recommendation, resume classification, and
formatting, layout, and language.
information extraction involves several key steps. This
Pokhare [5] outlines an approach that employs NLP and methodology integrates Natural Language Processing
(NLP) techniques and machine learning algorithms to
machine learning to parse, extract, and summarize data
automate the resume screening process, enhancing
from PDF resumes. This system identifies key sections efficiency and accuracy.
such as contact details, education, and work experience,
and uses NLP and ML techniques to extract and analyze i. Data Collection
keywords, skills, and qualifications. The extracted data is
stored for comparison and further analysis. Acquire a diverse and representative dataset of resumes
to ensure the system can handle various job profiles and
Kelkar and team [6] propose a Company Recommender categories.
System designed to match candidates to the best-fit
company. Their system uses text mining and machine ➢ Sources: Gather resumes from job boards, company
learning to rank resumes based on company-specific career pages, and publicly available datasets.
requirements. By extracting relevant information from Ensuring a diverse set of sources helps capture
resumes and comparing it to job specifications, the different resume styles and formats.
system assigns scores and leverages a machine learning ➢ Dataset Composition: The dataset should include key
model trained on historical data to identify patterns. This fields such as resume text, job category, skills,
approach ranks resumes and provides a list of education, and work experience. This diversity will
recommended candidates to recruiters. support comprehensive analysis and effective
classification.
Sinha and colleagues [7] introduce a method for resume
screening utilizing NLP and ML algorithms. The goal is ii. Data Preprocessing
to automate the CV screening process, allowing the
system to analyze and extract essential information from Prepare the collected resume data for analysis by cleaning
unstructured text. and standardizing it.

Shubham Bhor and his team [8] propose a solution for ➢ Text Cleaning: Remove unnecessary elements such
resume parsing using NLP techniques. Their goal is to as headers, footers, and irrelevant text to focus on the
streamline the identification of suitable candidates for job core resume content. Normalize text by converting it
openings by extracting key details from resumes to lowercase to maintain consistency.
uploaded to job portals. ➢ Tokenization: Break down the text into individual
words or tokens. This step simplifies the text for
Sanjana and her team [9] focus on resume validation and
further analysis.
filtering using built-in NLP techniques. Their approach
includes checking the accuracy of resumes and filtering

© 2024, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM37514 | Page 2

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 08 Issue: 09 | Sept - 2024 SJIF Rating: 8.448 ISSN: 2582-3930

➢ Lemmatization/Stemming: Reduce words to their ➢ Similarity Matching: Implement similarity measures,

base or root forms (e.g., "running" to "run"). This such as cosine similarity, to compare features
standardization helps in consolidating similar terms. extracted from resumes with job descriptions. This
➢ Vectorization: Convert the cleaned and tokenized comparison helps in identifying relevant job matches.
text into numerical representations using techniques ➢ Ranking Mechanism: Develop a ranking system to
such as Term Frequency-Inverse Document order job recommendations based on their relevance
Frequency (TF-IDF) or Bag-of-Words. This scores. The scores are derived from the similarity
transformation makes the text data usable for measures to prioritize the most suitable job
machine learning models. opportunities.

iii. Feature Extraction vi. Performance Evaluation

Extract and identify key features from resumes that are Assess the overall effectiveness of the Smart Resume
necessary for effective classification and Analyzer and its components.
recommendation.
➢ Cross-validation: Conduct k-fold cross-validation to
➢ Skills Extraction: Use NLP methods to identify and ensure that the model generalizes well across
extract specific skills mentioned in the resumes. different subsets of the data. This process helps in
Techniques such as Named Entity Recognition verifying the robustness of the model.
(NER) can help in detecting and categorizing these ➢ Confusion Matrix Analysis: Analyze the confusion
skills. matrix to understand the distribution of true positives,
➢ Experience and Education Parsing: Extract structured false positives, true negatives, and false negatives.
information about candidates' work experience and
educational background. This includes identifying
job titles, company names, roles, durations, and
academic degrees.

iv. Resume Classification

Develop and train a classification model to categorize

resumes into predefined job profiles or categories.

➢ Model Selection: Implement various machine

learning algorithms to classify resumes. Options
include Naïve Bayes, Support Vector Machines
(SVM), Random Forests, and Gradient Boosting.
Consider modern approaches such as transformer-
based models like BERT for enhanced accuracy.
➢ Training and Validation: Split the dataset into
training and validation sets. Train the model using the
training data and assess its performance on the
validation set. Evaluate the model using metrics such
as accuracy, precision, recall, and F1-score to ensure
it performs effectively.

v. Job Recommendation System

Fig -1: Flow Chart Diagram
Integrate a recommendation engine to suggest suitable
job positions based on the classified resumes.

© 2024, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM37514 | Page 3

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 08 Issue: 09 | Sept - 2024 SJIF Rating: 8.448 ISSN: 2582-3930

4. MODEL WITH EXPERIMENTAL RESULT job recommendation generation. The parsing module
manages a wide range of resume formats well, though
The experimental results of the Smart Resume improvements are needed to better handle more complex
Analyzer highlight its capability to efficiently process designs. Meanwhile, the recommendation system,
and analyze PDF resumes, subsequently offering relevant particularly the SVM model, provides exceptional
job recommendations. The PDF parsing module of the accuracy in matching resumes with job descriptions.
system shows notable proficiency in extracting and Ongoing refinement of both components will likely
standardizing text from resumes. This functionality is further boost the overall effectiveness of the Smart
crucial as it translates a variety of complex resume Resume Analyzer, solidifying its role as a valuable tool
formats into a uniform structure that can be evaluated for job seekers.
systematically. However, the system does face challenges
when dealing with more elaborate resume designs, such Fig -2: Category & Count Plotting for Classification
as those with intricate layouts, unconventional fonts, and
embedded graphics. These elements can complicate text
extraction and normalization, indicating a need for further
enhancement of the parsing algorithms to handle such
variations more effectively.
Nevertheless, the overall performance of the PDF parsing
module remains robust. It effectively processes a broad
spectrum of resume formats, converting them into a
format suitable for the recommendation system. This
transformation is essential for the recommendation
algorithms to operate efficiently. Consequently, users can
generally trust the Smart Resume Analyzer to handle
most resumes adeptly, though those with particularly
sophisticated designs might encounter some limitations.

In terms of job recommendations, the system’s

recommendation engine shows significant efficacy in
aligning resumes with appropriate job descriptions. This
aspect is central to the system’s value, as it ensures job
seekers receive suggestions that match their
qualifications and career goals. The success of the
recommendation system is largely due to the Support
Vector Machine (SVM) model, which outperforms the
Logistic Regression model in terms of accuracy.
The SVM model excels because it effectively manages
high-dimensional data and constructs a decision Fig -3: Model Accuracy Evaluation
boundary that maximizes the margin between distinct
classes. This attribute makes SVM particularly suitable
for complex classification tasks, such as matching
resumes with job descriptions that may contain nuanced
requirements. The SVM model’s ability to capture these
subtleties results in more accurate and relevant job
recommendations for users.

Conversely, the Logistic Regression model, while

effective for many classification tasks, falls short in this
scenario. Logistic Regression generally relies on a linear
decision boundary, which may not sufficiently address
the complexities involved in resume and job description
matching. Consequently, the superior performance of the
SVM model offers a notable advantage for the Smart
Resume Analyzer, enhancing its effectiveness in
providing tailored job recommendations. Fig -4: Confusion Matrix

The Smart Resume Analyzer demonstrates

significant proficiency in both PDF resume parsing and

© 2024, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM37514 | Page 4

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 08 Issue: 09 | Sept - 2024 SJIF Rating: 8.448 ISSN: 2582-3930

streamline the recruitment process by automatically

evaluating and ranking resumes, enabling HR teams to
quickly identify the most suitable candidates for a given
role. This review delves into the methodologies and
technologies that underpin AI-driven resume analyzers,
highlighting their potential to transform traditional hiring
practices.

Smart resume analyzers leverage natural language

processing to intelligently extract and align keywords,
Fig -5: Resume Category Prediction skills, and qualifications from resumes with the specific
criteria outlined in job descriptions. By analyzing
language patterns and contextual relevance, these tools
ensure a precise match between a candidate’s profile and
the job requirements, streamlining the hiring process and
enhancing the accuracy of candidate selection.

REFERENCES

[1] NLP based Extraction of Relevant Resume using

Machine Learning. Author: Nirali Bhaliya, Jay
Gandhi, Dheeraj Kumar Singh. Published in
International Journal of Innovative Technology and
Exploring Engineering (IJITEE) ISSN: 2278-3075
(Online), Volume-9 Issue-7, May 2020.\

[2] A Review Paper on Resume scanning using python.

Fig -6: Job Recommendation Author: Panil Jain, Nandkishor Kamble, Risha
Nadar, Shaikh Luqman. Published in International
Research Journal of Engineering and Technology
(IRJET), Volume: 09 Issue: 04 — Apr 2022.

[3] Resume Recommendation using Machine Learning.

Author: Ramba S Naik, Shrinivas R Dhotre.
Published in International Journal of Creative
Research Thoughts (IJCRT), Volume 10, Issue 7 July
2022.

[4] Resume Screening Using LSTM. Author: Navale

Sakshi, Doke Samiksha, Mule Divya3, Prof. Said S.
K. Published in International Journal of Research
Publication and Reviews, April 2022.

[5] “The study on the effectiveness of talent acquisition

Fig -7: Result through E-recruitment" by Prof. Sahana G N and
Pallav N. Published in the International Journal of
5. CONCLUSION Research and Analytical Reviews. May 2023.

In this paper, we reviewed the processes of Resume [6] “A review of machine learning applications in human
resource management" by Swati Garg, Shuchi Sinha,
Screening and Shortlisting, focusing on the role of AI- Arpan Kumar Kar, and Mauricio Mani. Published in
powered resume analyzers. These advanced tools are the International Journal of Productivity and
crafted to support HR professionals by significantly Performance Management. February 2021.
enhancing the efficiency of candidate screening and
shortlisting. Through the use of AI, these analyzers

© 2024, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM37514 | Page 5