Vector Database-Backed RAG for Enterprise HR Analytics

This study addresses the inefficiencies of the manual faculty promotion evaluation process at Mindanao State University–Maigo College of Education, Science and Technology (MSU-MCEST), which follows the 2005 Revised Integrated Scheme for Ranking and Promotion (ISRP). The traditional paper-based approach is time-consuming, prone to human error, and requires extensive administrative effort. To address these challenges, the study developed an automated decision-support system using a Vector Database–Backed Retrieval-Augmented Generation (RAG) framework. The system integrates Optical Character Recognition (OCR), Natural Language Processing (NLP), and semantic vector embeddings to transform unstructured 201 files into structured evaluation reports. A service-oriented architecture was implemented using XAMPP for the web interface and Python FastAPI for machine learning services, with ChromaDB enabling efficient similarity search and retrieval. Evaluation using 100 faculty records (700 document pages) achieved a classification accuracy of 97.14% (F1 = 0.966) and reduced processing time from three days to four hours. Statistical analysis showed no significant difference between automated and manual scoring (p > 0.05). ISO 25010 evaluation results indicated high system acceptability (Mean = 3.653). The findings demonstrate that the proposed system improves efficiency, accuracy, and transparency in faculty promotion pre-evaluation while maintaining compliance with institutional policies.

Retrieval-Augmented Generation (RAG)

1. M. C. Bacasong and R. D. Dinawanao, “Decision support system for academic ranking using Prolog logic programming,” International Journal of Information Technology and Computer Science, vol. 11, no. 3, pp. 45–52, 2019. [Google Scholar] [Crossref]

2. A. L. Cabaobao and P. S. Malubag, “Workflow automation in government agencies: A case study,” Philippine Journal of Public Administration, vol. 64, no. 2, pp. 78–95, 2020. [Google Scholar] [Crossref]

3. Commission on Higher Education, “A.C.H.I.E.V.E. 2030: Strategic roadmap for Philippine higher education,” CHED Memorandum Order No. 2020-001, 2020. [Google Scholar] [Crossref]

4. S. Es, A. Bhattacharjee, and D. Varshney, “RAGAS: Automated evaluation of retrieval-augmented generation,” arXiv preprint arXiv:2309.15217, 2023. [Google Scholar] [Crossref]

5. Y. Gao et al., “Retrieval-augmented generation for large language models: A survey,” arXiv preprint arXiv:2312.10997, 2023. [Google Scholar] [Crossref]

6. J. He, S. Zhang, and Y. Li, “Optimization challenges in vector indexing for real-time applications,” IEEE Transactions on Knowledge and Data Engineering, vol. 34, no. 8, pp. 3721–3734, 2022. [Google Scholar] [Crossref]

7. P. Lewis et al., “Retrieval-augmented generation for knowledge-intensive NLP tasks,” Advances in Neural Information Processing Systems, vol. 33, pp. 9459–9474, 2020. [Google Scholar] [Crossref]

8. J. A. Lucero et al., “E-learning readiness among Philippine higher education institutions,” Asian Journal of Distance Education, vol. 16, no. 1, pp. 123–145, 2021. [Google Scholar] [Crossref]

9. Y. A. Malkov and D. A. Yashunin, “Efficient and robust approximate nearest neighbor search using hierarchical navigable small world graphs,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 42, no. 4, pp. 824–836, 2020. [Google Scholar] [Crossref]

10. T. T. Nguyen et al., “RAG-driven analytics for HR decision support,” Journal of Human Resource Management, vol. 28, no. 4, pp. 445–462, 2023. [Google Scholar] [Crossref]

11. J. Pan et al., “A survey on vector database management systems,” The VLDB Journal, vol. 32, no. 5, pp. 1041–1065, 2023. [Google Scholar] [Crossref]

12. N. Reimers and I. Gurevych, “Sentence-BERT: Sentence embeddings using Siamese BERT-networks,” in Proc. EMNLP, 2019, pp. 3982–3992. [Google Scholar] [Crossref]

13. W. Shi et al., “Adaptive RAG systems for document classification,” ACM Transactions on Information Systems, vol. 41, no. 3, pp. 1–28, 2023. [Google Scholar] [Crossref]

14. H. Touvron et al., “Llama 2: Open foundation and fine-tuned chat models,” arXiv preprint arXiv:2307.09288, 2023. [Google Scholar] [Crossref]

15. X. Wang et al., “Semantic search implementations using vector databases,” Information Processing & Management, vol. 59, no. 6, 2022. [Google Scholar] [Crossref]

• What the Desert Fathers Teach Data Scientists: Ancient Ascetic Principles for Ethical Machine-Learning Practice
• Comparative Analysis of Some Machine Learning Algorithms for the Classification of Ransomware
• Comparative Performance Analysis of Some Priority Queue Variants in Dijkstra’s Algorithm
• Transfer Learning in Detecting E-Assessment Malpractice from a Proctored Video Recordings.
• Dual-Modal Detection of Parkinson’s Disease: A Clinical Framework and Deep Learning Approach Using NeuroParkNet