<p>Ant-Miner, a rule-based classifier, has been extensively utilized for classification tasks. However, it features numerous controlling parameters that significantly impact its performance. The standard Ant-Miner (AM) often encounters issues such as slow or premature convergence and high selective pressure. Several studies have proposed innovative methods to enhance AM’s performance by examining the quality function, heuristic strategies, and pheromone update mechanisms. Recent research has also focused on ant selection for pheromone updates and term selection strategies. Despite substantial efforts, existing studies are limited by challenges like local optima entrapment, premature or slow convergence, high selective pressure, exhaustive searches, and the generation of specific yet low-quality rules. This paper introduces a novel Ant-Miner algorithm (ASAQ—AM) that implements novel Ant-Selection strategies for pheromone updating and data classification, alongside an Advanced Quality function. This novel quality function employs hierarchical thresholds to evaluate ants based on coverage, rule length, and accuracy. The selected ants contribute to pheromone updates on their respective paths guided by the proposed quality function. The effectiveness of the ASAQ—AM approach is assessed using four publicly available datasets and standard benchmark performance metrics, including accuracy and F1-score. The results demonstrate that the proposed method outperforms the basic Ant-Miner, state-of-the-art variants, and several data mining approaches in terms of accuracy, F1-score, and convergence speed.</p>