Abstract :

Online Job Advertisements Contain Large Volumes Of Unstructured Textual Data That Make Automatic Job Title Identification A Challenging Task. Traditional Classification Approaches Require Large Labeled Datasets And Often Fail To Capture Semantic Relationships Between Job Descriptions And Occupation Titles. To Address These Limitations, This Paper Presents A Two Stage Job Title Identification System For Online Job Advertisements Using Machine Learning And Natural Language Processing Techniques. The Proposed System Combines Supervised And Unsupervised Learning Methods To Improve Job Title Prediction Accuracy With Minimal Labeled Data. In The First Stage, Bidirectional Encoder Representations From Transformers (BERT) Is Used To Classify Job Advertisements Into Their Corresponding Sectors Such As Information Technology, Agriculture, And Sales. In The Second Stage, Document Embedding And Similarity Matching Techniques Are Applied To Identify The Most Relevant Occupation Title From The Predicted Sector. The System Utilizes Feature Extraction, Document Representation, And Cosine Similarity Measures To Improve Semantic Understanding Of Job Descriptions. Experimental Evaluation Demonstrates That The Proposed Methodology Significantly Improves Job Title Identification Accuracy Compared With Traditional Machine Learning Approaches Such As Support Vector Machine (SVM), Naïve Bayes, And Logistic Regression. The Proposed Framework Is Scalable, Efficient, And Adaptable To Multilingual Job Market Datasets. Furthermore, The System Can Support Recruitment Analytics, Labor Market Analysis, And Career Guidance Applications By Identifying Emerging Occupations And High-demand Job Roles From Online Recruitment Platforms.

Published:

20-5-2026

Issue:

Vol. 26 No. 5 (2026)

Page Nos:

1268-1273

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite