Jyothsna Vamisetti, Gummadidala Sai Krishna, Gutha Chaitanya, Kuruva Bharath Kumar, Muriki Tarun

DOI: https://doi.org/10.5281/zenodo.22130222

The recruitment landscape is undergoing rapid transformation as organisations increasingly rely on artificial intelligence to manage high applicant volumes, standardise candidate evaluation, and reduce hiring time and cost. Existing AI-based career tools typically address only a single stage of this process — screening resumes, running generic interview chatbots, or flagging fraudulent postings in isolation — rather than supporting a candidate end to end. This paper presents CareerMap, an integrated AI-driven framework that unifies five interconnected capabilities within one human-centred pipeline: automated resume parsing and entity extraction, embedding-based skill-gap analysis, a conversational mock-interview engine, fraud and inconsistency detection for job postings and candidate claims, and affect-aware feedback derived from behavioural and emotional signals. CareerMap combines transformer-based language models (BERT and its derivatives) with BiLSTM-CRF sequence labelling for entity recognition, cosine similarity for skill matching, a weighted fraud-scoring model, and speech- and text-based affective computing for emotion recognition. To situate CareerMap within existing work, this paper conducts a comparative review of thirty-six prior studies spanning named entity recognition, skill embeddings, explainable AI, deceptive-language detection, and bias in machine learning, showing that no existing system integrates these techniques into a single coherent pipeline. Building on this gap, the proposed methodology is described in detail, including the hybrid BERT-BiLSTM-CRF resume-parsing pipeline, the cosine-similarity formulation for skill-gap ranking, and the weighted fraud-scoring metric. Drawing on performance patterns reported across the reviewed literature, a comparative analysis across all five functional modules indicates that transformer-based contextual models consistently outperform classical rule-based and shallow statistical baselines in information-extraction accuracy, explanation quality, and user engagement, a pattern that directly motivates CareerMap’s architectural choices. The paper concludes that an integrated, explainable, emotionally aware approach to career preparation offers a more trustworthy alternative to fragmented point-solutions, and outlines future work including longitudinal validation with real hiring outcomes and multilingual expansion of the fraud-detection module.

Resume Parsing, Natural Language Processing, BERT-based Transformer Models, Skill Gap Detection, Job Recommendation, Mock-Interview Simulation, Emotion Recognition, Explainable AI, Fraud Detection.

This study has examined how a range of classical machine-learning models and modern transformer-based deep learning models perform across the core tasks that underpin AI-assisted recruitment systems. Consistently across every module evaluated — resume parsing, skill matching, fraud detection, interview scoring, and emotion recognition — models that incorporate transformer-based contextual representations, exemplified by BERT and its derivatives, outperformed traditional rule-based and shallow statistical methods. In resume parsing specifically, contextual transformer models proved substantially better at identifying relationships between words in unstructured text than approaches that rely on fixed rules or purely local features. In skill matching, representing words in context outperformed representing them as fixed, context-independent vectors. Fraud-detection models were most effective when they combined multiple complementary signals and paired this with explainable outputs, which improved both accuracy and the transparency needed to support human review. Interview-scoring systems performed better when they incorporated behavioural indicators alongside semantic content, bringing automated scores closer to the judgement of human evaluators. Finally, emotion-recognition models demonstrated that combining multiple data modalities, such as audio and video, improves the reliability with which stress, confidence, and other affective states can be detected during mock interviews. These findings, taken as a whole, support the central argument of this paper: that an integrated, explainable, and emotionally aware approach to career preparation offers a more complete and more trustworthy alternative to the fragmented, single-purpose tools that currently dominate the market. Future work will focus on validating CareerMap against real-world hiring outcomes over a longitudinal study, extending the fraud-detection module to handle multilingual and cross-regional job postings, incorporating additional fairness constraints to further mitigate bias across demographic groups, and conducting user studies to measure the framework’s impact on candidate confidence and interview performance over time.

[1] J. Lafferty, A. McCallum, and F. Pereira, "Conditional Random Fields: Probabilistic Models for Segmenting and Labeling
Sequence Data," in Proc. 18th Int. Conf. Machine Learning (ICML), 2001, pp. 282–289.
[2] Z. Huang, W. Xu, and K. Yu, "Bidirectional LSTM-CRF Models for Sequence Tagging," arXiv preprint arXiv:1508.01991,
2015.
[3] J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, "BERT: Pre-training of Deep Bidirectional Transformers for Language
Understanding," in Proc. NAACL-HLT, 2019, pp. 4171–4186.
[4] T. Mikolov, K. Chen, G. Corrado, and J. Dean, "Efficient Estimation of Word Representations in Vector Space," in Proc.
NIPS Workshop, 2013.
[5] V. Pérez-Rosas, B. Kleinberg, A. Lefevre, and R. Mihalcea, "Automatic Detection of Deceptive Language," in Proc. ACL,
2015.
[6] B. W. Schuller et al., "Speech Emotion Recognition: Two Decades in a Nutshell, Benchmarks, and Ongoing Trends," IEEE
Signal Processing Magazine, vol. 35, no. 4, pp. 154–159, 2018.
[7] M. T. Ribeiro, S. Singh, and C. Guestrin, "’Why Should I Trust You?’: Explaining the Predictions of Any Classifier," in
Proc. ACM SIGKDD, 2016.
[8] S. M. Lundberg and S.-I. Lee, "A Unified Approach to Interpreting Model Predictions," in Proc. NeurIPS, 2017.
[9] E. F. Tjong Kim Sang and F. De Meulder, "Introduction to the CoNLL-2003 Shared Task: Language-Independent Named
Entity Recognition," in Proc. CoNLL, 2003.
[10] M. E. Peters et al., "Deep Contextualized Word Representations," in Proc. NAACL-HLT, 2018.
[11] A. Vaswani et al., "Attention Is All You Need," in Proc. NeurIPS, 2017.
[12] J. Pennington, R. Socher, and C. Manning, "GloVe: Global Vectors for Word Representation," in Proc. EMNLP, 2014.
[13] D. Jurafsky and J. H. Martin, Speech and Language Processing, 3rd ed. Draft, 2023.
[14] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. Cambridge, MA, USA: MIT Press, 2016.
[15] S. Hochreiter and J. Schmidhuber, "Long Short-Term Memory," Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997.
[16] Y. Kim, "Convolutional Neural Networks for Sentence Classification," in Proc. EMNLP, 2014.
[17] Y. Liu et al., "RoBERTa: A Robustly Optimized BERT Pretraining Approach," arXiv preprint arXiv:1907.11692, 2019.
[18] A. Radford et al., "Improving Language Understanding by Generative Pre-Training," OpenAI, 2018.
[19] S. Mayhew, C. Tsai, and D. Roth, "Fine-Grained Entity Typing with High-Multiplicity Assignments," in Proc. ACL,
2017.
[20] A.-S. Dadzie and M. A. Rowe, "Approaches to Skill Ontologies and Career Knowledge Representation: A Survey," 2019.
[21] E. Strubell, P. Verga, D. Belanger, and A. McCallum, "Fast and Accurate Entity Recognition with Iterated Dilated
Convolutions," in Proc. EMNLP, 2017.
[22] M. Schmitt et al., "Automated Interview Scoring Using Deep Learning Techniques," 2020.
[23] A. Caliskan, J. J. Bryson, and A. Narayanan, "Semantics Derived Automatically from Language Corpora Contain Human
Like Biases," Science, vol. 356, no. 6334, pp. 183–186, 2017.
[24] N. Mehrabi et al., "A Survey on Bias and Fairness in Machine Learning," ACM Computing Surveys, vol. 54, no. 6, 2021.
[25] L. Breiman, "Random Forests," Machine Learning, vol. 45, no. 1, pp. 5–32, 2001.
[26] C. Cortes and V. Vapnik, "Support-Vector Networks," Machine Learning, vol. 20, no. 3, pp. 273–297, 1995.
[27] C. M. Bishop, Pattern Recognition and Machine Learning. New York, NY, USA: Springer, 2006.
[28] Organisation for Economic Co-operation and Development (OECD), OECD Principles on Artificial Intelligence, 2019.
[29] V. Sanh, L. Debut, J. Chaumond, and T. Wolf, "DistilBERT, a Distilled Version of BERT: Smaller, Faster, Cheaper and
Lighter," arXiv preprint arXiv:1910.01108, 2019.
[30] Z. Yang, Z. Dai, Y. Yang, J. Carbonell, R. Salakhutdinov, and Q. V. Le, "XLNet: Generalized Autoregressive Pretraining
for Language Understanding," in Proc. NeurIPS, 2019, pp. 5754–5764.
[31] M. le Vrang, A. Papantoniou, E. Pauwels, P. Fannes, D. Vandensteen, and J. De Smedt, "ESCO: Boosting Job Matching
in Europe with Semantic Interoperability," Computer, vol. 47, no. 10, pp. 57–64, 2014.
[32] M. N. Freire and L. N. de Castro, "e-Recruitment Recommender Systems: A Systematic Review," Knowledge and
Information Systems, vol. 63, no. 1, pp. 1–20, 2021.
[33] S. Vidros, C. Kolias, G. Kambourakis, and L. Akoglu, "Automatic Detection of Online Recruitment Frauds:
Characteristics, Methods, and a Public Dataset," Future Internet, vol. 9, no. 1, p. 6, 2017.
[34] T. Baltrušaitis, C. Ahuja, and L.-P. Morency, "Multimodal Machine Learning: A Survey and Taxonomy," IEEE Trans.
Pattern Analysis and Machine Intelligence, vol. 41, no. 2, pp. 423–443, 2019.
[35] M. Raghavan, S. Barocas, J. Kleinberg, and K. Levy, "Mitigating Bias in Algorithmic Hiring: Evaluating Claims and
Practices," in Proc. ACM Conf. Fairness, Accountability, and Transparency (FAT*), 2020, pp. 469–481.
[36] I. Naim, M. I. Tanveer, D. Gildea, and M. E. Hoque, "Automated Analysis and Prediction of Job Interview Performance,"
IEEE Transactions on Affective Computing, vol. 9, no. 2, pp. 191–204, 2018.

Jyothsna Vamisetti, Gummadidala Sai Krishna, Gutha Chaitanya, Kuruva Bharath Kumar, Muriki Tarun (2026). CareerMap: A Comparative AI Framework for Resume Parsing, Skill Gap Analysis, and Intelligent Interview Assistance. International Journal of Computer Science Engineering Techniques, 10(4), 100–113. ISSN: 2455-135X. DOI: https://doi.org/10.5281/zenodo.22130222