“…In terms of Accuracy, our proposed approach outperformed the methods used in [29], [47], [10], [7], [24], [26], [25], [11] by margins of 3.7%, 0.7%, 2.7%, 1.2%, 2.2%, 2.6%, 1.2%, and 1.2%, respectively. When compared to the Precision, the methods used in [10], [7], [24], [26], [25], [11] were outperformed by 2.4%, 0.2%, 0.9%, 3.2%, 1.8%, and 0.9%, respectively. When compared to the Recall, the methods used in [47], [10], [7], [24], [26], [25], [11] were outperformed by 1.4%, 2.9%, 1.3%, 3.5%, 1.9%, 0.4%, and 1.4%, respectively.…”