Home / Current Issue / Paper 1717183
Deep Semantic Matching for Intelligent Resume Screening Using Hybrid Transformer Architectures
Subject area: Science,Engineering and Technology · Area of research: Deep Learning
DOI: https://doi.org/10.64388/IREV9I11-1717183
Abstract
With the fast growth of digital recruitment sources, the quantity of employment applications has grown to an all-time high, and screening resumes manually is no longer efficient and effective. The traditional automated screening tools mostly use the method of matching keywords, which does not achieve the semantic background, transferable competencies, and subtle match between applicant profile and the job requirements. To overcome these shortcomings, this paper suggests a smart resume screening system on the deep semantic matching on the basis of hybrid transformer frameworks. The proposed solution involves transformer encoders (specialized to resume and job description) along with cross-attention and feature fusion to predict fine-grained contextual relation. The system can score relevance accurately across beyond the surface-level similarity of text and therefore project both documents into common semantic embedding space. Further semantic skill normalization and experience-sensitive weighting enhance the robustness of the skills in a wide range of resume forms and domains. The experimental analysis based on publicly available data proves that the suggested framework is more accurate and more effective in ranking compared to key-powered and single-transformer baselines. The findings underscore the scalability and interpretation and next-generation recruitment systems possibilities of the framework.
Keywords
Intelligent Resume Screening, Deep Semantic Matching, Hybrid Transformer Architecture, Resume–Job Matching, Natural Language Processing
How to cite this paper
@article{1717183,
author = {Nirmala Priyadharshini P, Devaguru M, Jeyaprakash J, Abinass M, Sanjai N},
title = {Deep Semantic Matching for Intelligent Resume Screening Using Hybrid Transformer Architectures},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {9},
number = {11},
pages = {568-575},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1717183.pdf},
abstract = {With the fast growth of digital recruitment sources, the quantity of employment applications has grown to an all-time high, and screening resumes manually is no longer efficient and effective. The traditional automated screening tools mostly use the method of matching keywords, which does not achieve the semantic background, transferable competencies, and subtle match between applicant profile and the job requirements. To overcome these shortcomings, this paper suggests a smart resume screening system on the deep semantic matching on the basis of hybrid transformer frameworks. The proposed solution involves transformer encoders (specialized to resume and job description) along with cross-attention and feature fusion to predict fine-grained contextual relation. The system can score relevance accurately across beyond the surface-level similarity of text and therefore project both documents into common semantic embedding space. Further semantic skill normalization and experience-sensitive weighting enhance the robustness of the skills in a wide range of resume forms and domains. The experimental analysis based on publicly available data proves that the suggested framework is more accurate and more effective in ranking compared to key-powered and single-transformer baselines. The findings underscore the scalability and interpretation and next-generation recruitment systems possibilities of the framework.},
keywords = {Intelligent Resume Screening, Deep Semantic Matching, Hybrid Transformer Architecture, Resume–Job Matching, Natural Language Processing},
month = {May},
doi = {https://doi.org/10.64388/IREV9I11-1717183}
}