Semantic Retrieval and Knowledge Graph Integration for Intelligent Urban Document Systems: A Hybrid Machine Learning Framework for Smart Municipal Governance

Efficient access to municipal documents is a critical requirement for smart urban governance. Traditional keyword-based retrieval systems fail to capture semantic relationships among heterogeneous urban documents, leading to inefficient information access and weak knowledge reuse across departments. This study proposes a Hybrid Semantic Retrieval and Knowledge Graph Integration Framework (HSR-KG) to enhance intelligent document management in municipal environments. The framework integrates Transformer-based semantic embeddings with a Knowledge Graph structure to enable context-aware retrieval and relationship-aware knowledge discovery. The proposed system is evaluated using real-world municipal documents from the Shiraz Municipality. Experimental results demonstrate that the proposed model significantly outperforms traditional keyword-based retrieval methods in terms of Precision@K, Recall @K, and Mean Reciprocal Rank (MRR). Specifically, the hybrid model achieves a Precision @ ۱۰ of ۰.۸۸, compared to ۰.۷۱ in keyword-based systems, indicating a substantial improvement in retrieval accuracy and semantic understanding.

کلیدواژه ها:

Smart City ، Semantic Retrieval ، Knowledge Graph ، NLP ، Urban Computing ، Machine Learning ، Information Systems