AI-Based Resume Parser with ATS Scoring

This paper presented an AI-Based Resume Parser with ATS Scoring, a practical web application that automates the evalu- ation of candidate resumes against job descriptions using rule- based NLP and TF-IDF similarity scoring. The system extracts text from PDF and DOCX files, applies SpaCy-based lemma- tization and stop-word removal, computes a three-phase ATS compatibility score combining cosine similarity with skill- keyword boosting, and audits resume structure against eight quality heuristics. Missing keywords are identified through lemmatized set subtraction and returned to the candidate in ranked order. A locally hosted language model generates role- specific coaching tips and resume rewrites while keeping candidate data within the institutional environment. The system is implemented as a Flask web application with role-differentiated student and recruiter dashboards, SQLite persistence, and session-based authentication. All scoring logic is deterministic and auditable, with no trained machine learn- ing classifier in the current implementation. The system fulfills its core design objective: providing candidates with a transparent, actionable breakdown of their resume’s alignment with a specific job description, and pro- viding recruiters with a ranked applicant list derived from an interpretable scoring process.

[1]     K. Khelkhal and D. Lanasri, “Smart-Hiring: An Explainable End-to- End Pipeline for CV Information Extraction and Job Matching,” arXiv preprint arXiv:2511.02537, 2025. [2]     A. Akkasi, “Job Description Parsing with Explainable Transformer- Based Skill Extraction,” Expert Systems with Applications, vol. 240, 2024. [3]     F. Retyk, H. Fabregat, J. Aizpuru, M. Taglio, and R. Zbib, “Re´sume´ Parsing as Hierarchical Sequence Labeling,” in Proc. EMNLP Workshop on Document Intelligence, 2022, pp. 45–54. [4]     S. Patil and A. Kulkarni, “Resume Screening Automation Using Natural Language Processing Techniques,” International Journal of Innovative Science and Research Technology, vol. 8, no. 3, pp. 1023–1031, 2023. [5]     S. Karhale, P. Pawade, P. Sukala, S. Swami, and A. Khan, “Resume Analysis System Using Natural Language Processing,” International Journal of Latest Technology in Engineering, Management & Applied Science, vol. 14, no. 2, 2025. [6]     R. V. K. Bevara, N. R. Mannuru, S. P. Karedla, B. D. Lund, T. Xiao, and H. Pasem, “Resume2Vec: Transformer-Based Embeddings for Candidate-Job Compatibility,” in Proc. IEEE ICMLA, 2024, pp. 312– 319. [7]     N. Reimers and I. Gurevych, “Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks,” in Proc. EMNLP-IJCNLP, Hong Kong, 2019, pp. 3982–3992. [8]     J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” in Proc. NAACL-HLT, Minneapolis, MN, 2019, pp. 4171–4186. [9]     Z. Chen, “Ethics and Discrimination in Artificial Intelligence-Enabled Recruitment Practices,” Humanities and Social Sciences Communica- tions, vol. 10, no. 1, p. 567, 2023. [10]    A. Fabris, “Fairness and Bias in Algorithmic Hiring: A Multidisciplinary Survey,” ACM Computing Surveys, vol. 57, no. 4, 2025. [11]    M. Langenkamp, A. Costa, and C. Cheung, “Hiring Fairly in the Age of Algorithms,” arXiv preprint arXiv:2004.07132, 2020.