Table 4.

Result comparison of TextConvoNet with attention and/or transformer-based deep learning models

	Accuracy	Precision	Recall	F1-score	Specificity	Gmean 1	Gmean 2	MCC
DATASET-1
BiLSTM + Attention	0.806	0.816	0.790	0.803	0.822	0.803	0.806	0.613
BERT	0.776	0.736	0.859	0.793	0.694	0.795	0.772	0.560
HAN	0.802	0.791	0.821	0.806	0.783	0.806	0.802	0.606
BerConvoNet	0.831	0.802	0.797	0.794	0.697	0.799	0.814	0.640
CNN-BiLSTM	0.820	0.796	0.859	0.826	0.781	0.827	0.819	0.642
TLGNN	0.710	0.725	0.675	0.699	0.745	0.700	0.709	0.422
SeqGNN	0.821	0.787	0.882	0.828	0.763	0.829	0.820	0.649
TextConvoNet_6	0.822	0.848	0.783	0.814	0.860	0.815	0.821	0.645
TextConvoNet_4	0.819	0.833	0.798	0.815	0.841	0.815	0.819	0.639
DATASET-2
BiLSTM + Attention	0.886	0.883	0.846	0.864	0.915	0.864	0.880	0.766
BERT	0.772	0.685	0.867	0.765	0.700	0.771	0.779	0.564
HAN	0.863	0.880	0.788	0.831	0.919	0.833	0.851	0.720
BerConvoNet	0.883	0.821	0.911	0.866	0.859	0.849	0.852	0.755
CNN-BiLSTM	0.872	0.901	0.861	0.881	0.886	0.881	0.873	0.745
TLGNN	0.732	0.653	0.800	0.719	0.680	0.723	0.737	0.476
SeqGNN	0.821	0.780	0.882	0.828	0.763	0.829	0.820	0.649
TextConvoNet_6	0.904	0.905	0.867	0.886	0.932	0.886	0.899	0.804
TextConvoNet_4	0.872	0.819	0.902	0.858	0.849	0.859	0.875	0.745
DATASET-3
BiLSTM + Attention	0.990	0.960	0.960	0.960	0.994	0.960	0.977	0.954
BERT	0.956	0.825	0.825	0.825	0.975	0.825	0.897	0.801
HAN	0.890	0.560	0.560	0.560	0.937	0.560	0.724	0.497
BerConvoNet	0.992	0.968	0.968	0.968	0.995	0.966	0.982	0.966
CNN-BiLSTM	0.989	0.959	0.959	0.959	0.994	0.959	0.976	0.954
TLGNN	0.944	0.777	0.777	0.777	0.968	0.777	0.867	0.745
SeqGNN	0.967	0.962	0.819	0.885	0.994	0.887	0.902	0.870
TextConvoNet_6	0.992	0.968	0.968	0.968	0.995	0.968	0.981	0.963
TextConvoNet_4	0.992	0.969	0.969	0.969	0.996	0.969	0.982	0.965
DATASET-4
BiLSTM + Attention	0.861	0.792	0.792	0.792	0.896	0.792	0.843	0.689
BERT	0.808	0.712	0.712	0.712	0.856	0.712	0.7806	0.568
HAN	0.795	0.693	0.693	0.693	0.846	0.693	0.766	0.540
BerConvoNet	0.882	0.824	0.824	0.824	0.921	0.826	0.868	0.738
CNN-BiLSTM	0.841	0.762	0.762	0.762	0.881	0.762	0.819	0.643
TLGNN	0.800	0.700	0.706	0.706	0.850	0.706	0.771	0.551
SeqGNN	0.875	0.861	0.747	0.8	0.940	0.802	0.838	0.714
TextConvoNet_6	0.884	0.826	0.826	0.826	0.913	0.826	0.869	0.739
TextConvoNet_4	0.873	0.809	0.809	0.809	0.904	0.809	0.855	0.713
DATASET-5
BiLSTM + Attention	0.799	0.493	0.493	0.493	0.874	0.499	0.660	0.374
BERT	0.711	0.278	0.278	0.278	0.819	0.278	0.477	0.098
HAN	0.700	0.250	0.250	0.250	0.812	0.250	0.450	0.062
BerConvoNet	0.828	0.572	0.572	0.572	0.894	0.572	0.721	0.474
CNN-BiLSTM	0.830	0.576	0.576	0.576	0.89	0.576	0.718	0.470
TLGNN	0.729	0.322	0.322	0.322	0.830	0.322	0.517	0.153
SeqGNN	0.737	0.285	0.209	0.241	0.869	0.244	0.426	0.0885
TextConvoNet_6	0.839	0.597	0.597	0.597	0.899	0.597	0.732	0.496
TextConvoNet_4	0.828	0.571	0.571	0.571	0.893	0.571	0.714	0.463

Bold entries show the significant values