An Intelligent System for Flexible Access Control in Semantic Search Using Fuzzy Logic

Conventional access control systems in semantic search rely on crisp, binary logic (allow/deny), which often leads to the rejection of highly relevant results. This paper introduces an intelligent access control system based on Fuzzy Logic. The proposed Fuzzy Inference System (FIS) evaluates two inputs simultaneously: the semantic similarity of a document to a query and the access level difference. Based on a set of human-like rules, the system generates a flexible 'Permission Score.' Experimental results show that the fuzzy approach improves nuanced decision accuracy by over ۲۵% compared to a traditional crisp system, correctly identifying valuable information that would have otherwise been blocked. This work showcases a novel application of Soft Computing in information retrieval security.

کلیدواژه ها:

semantic search ، vector embeddings ، access control ، fuzzy logic ، soft computing