Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{193325,
        author = {Anil Mamidi},
        title = {A Vector-Based Representation of Human Skills Using Multimodal Data},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {16-21},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=193325},
        abstract = {Traditional representations of professional ability rely on resumes, job titles, and educational credentials, which provide an incomplete and often biased view of a person’s true skills. In this paper, we propose a multimodal representation learning framework that models human professional capability as a continuous vector embedding derived from heterogeneous data sources, including text, code, images, speech, and outcome-based performance signals. We introduce the Human Skill Vector (HSV), a unified latent representation constructed through a learnable fusion architecture with temporal weighting to prioritize recent evidence. Using a large-scale dataset built from publicly available professional artifacts, we demonstrate that HSV embeddings outperform resume-based and profile-based baselines in predicting job roles, performance metrics, and skill similarity. These results suggest that vector-based representations provide a more accurate and scalable foundation for talent discovery, hiring, and workforce analytics.},
        keywords = {Multimodal Representation Learning, Human Skill Representation, Professional Skill Embedding, Talent Intelligence, Machine Learning, Multimodal Data fusion, Neural Network.},
        month = {February},
        }