Enhancing Keyword Search in Relational Databases with Word Embeddings," IEEE Access, May 28, 2025

This article presents Key2Vec, a novel framework that applies Word2Vec word embeddings to improve keyword search in relational databases. The framework addresses challenges such as query ambiguity and complex structured relationships by semantically interpreting user keywords and optimizing query classes according to user intent and data needs. Key2Vec enhances both the relevance of search results and the efficiency of query processing. Experimental results on benchmark databases demonstrate an approximate 25% improvement in accuracy and a 30% reduction in query processing time compared to traditional keyword search methods. This research bridges natural language processing with database search technologies to provide a more intuitive and effective querying experience.