#Evaluation of Runs of the different Models ##1. Evaluation of Argument-Model ###1.1 Evaluation of NeuralNet-Argument Model with datasplit per image ####1.2.1 Evaluation over all labeled topics
| Run | P_Strong | P_Both |
|---|---|---|
| Run 1 | 0.5625 | 0.8325 |
| Run 2 | 0.5375 | 0.8525 |
| Run 3 | 0.5475 | 0.845 |
| Run 4 | 0.57 | 0.8425 |
| Run 5 | 0.585 | 0.8525 |
| Run 6 | 0.56 | 0.8575 |
| Run 7 | 0.5775 | 0.8575 |
| Run 8 | 0.48 | 0.8175 |
| Run 9 | 0.5725 | 0.85 |
| Run 10 | 0.58 | 0.865 |
| Average | 0.5572 | 0.8472 |
####1.2.2 Evaluation over all labeled valid topics
| Run | P_Strong | P_Both |
|---|---|---|
| Run 1 | 0.5944 | 0.8639 |
| Run 2 | 0.5583 | 0.8806 |
| Run 3 | 0.575 | 0.875 |
| Run 4 | 0.5944 | 0.8694 |
| Run 5 | 0.6194 | 0.8833 |
| Run 6 | 0.5944 | 0.8917 |
| Run 7 | 0.6083 | 0.8889 |
| Run 8 | 0.4972 | 0.8389 |
| Run 9 | 0.6 | 0.8778 |
| Run 10 | 0.6083 | 0.8944 |
| Average | 0.585 | 0.8764 |
###1.2 Evaluation of NeuralNet-Argument Model with datasplit per topic ####1.2.1 Evaluation over all labeled topics
| Run | P_Strong | P_Both |
|---|---|---|
| Run 1 | 0.55 | 0.8575 |
| Run 2 | 0.545 | 0.8425 |
| Run 3 | 0.5425 | 0.8575 |
| Run 4 | 0.5475 | 0.86 |
| Run 5 | 0.555 | 0.8475 |
| Run 6 | 0.5625 | 0.8925 |
| Run 7 | 0.5575 | 0.86 |
| Run 8 | 0.525 | 0.855 |
| Run 9 | 0.5725 | 0.88 |
| Run 10 | 0.4625 | 0.815 |
| Average | 0.542 | 0.8568 |
####1.2.2 Evaluation over all labeled valid topics
| Run | P_Strong | P_Both |
|---|---|---|
| Run 1 | 0.5861 | 0.8917 |
| Run 2 | 0.575 | 0.8722 |
| Run 3 | 0.5722 | 0.8806 |
| Run 4 | 0.5778 | 0.8861 |
| Run 5 | 0.575 | 0.8694 |
| Run 6 | 0.5917 | 0.9111 |
| Run 7 | 0.5806 | 0.8889 |
| Run 8 | 0.5444 | 0.8778 |
| Run 9 | 0.6028 | 0.9083 |
| Run 10 | 0.4861 | 0.8361 |
| Average | 0.5692 | 0.8822 |
####1.2.3 Evaluation over all labeled test topics
| Run | P_Strong | P_Both |
|---|---|---|
| Run 1 | 0.575 | 0.775 |
| Run 2 | 0.675 | 0.9625 |
| Run 3 | 0.4625 | 0.9 |
| Run 4 | 0.5167 | 0.8833 |
| Run 5 | 0.6 | 0.8833 |
| Run 6 | 0.4875 | 0.9125 |
| Run 7 | 0.5875 | 0.9 |
| Run 8 | 0.6125 | 0.8625 |
| Run 9 | 0.5 | 0.9333 |
| Run 10 | 0.5333 | 0.7167 |
| Average | 0.555 | 0.8729 |
##2. Evaluation of Stance-Model ###2.1 Evaluation of NeuralNet-Stance Model with datasplit per image ####2.1.1 Evaluation over all labeled topics
| Run | Accuracy |
|---|---|
| Run 1 | 0.4442 |
| Run 2 | 0.4797 |
| Run 3 | 0.485 |
| Run 4 | 0.4861 |
| Run 5 | 0.4272 |
| Run 6 | 0.4821 |
| Run 7 | 0.4855 |
| Run 8 | 0.4817 |
| Run 9 | 0.4841 |
| Run 10 | 0.467 |
| Average | 0.4723 |
####2.1.2 Evaluation over all labeled valid topics
| Run | Accuracy |
|---|---|
| Run 1 | 0.4645 |
| Run 2 | 0.508 |
| Run 3 | 0.5245 |
| Run 4 | 0.5118 |
| Run 5 | 0.4396 |
| Run 6 | 0.501 |
| Run 7 | 0.5174 |
| Run 8 | 0.493 |
| Run 9 | 0.5132 |
| Run 10 | 0.4879 |
| Average | 0.4961 |
###2.2 Evaluation of NeuralNet-Stance Model with datasplit per topic ####2.2.1 Evaluation over all labeled topics
| Run | Accuracy |
|---|---|
| Run 1 | 0.5167 |
| Run 2 | 0.4637 |
| Run 3 | 0.4834 |
| Run 4 | 0.4835 |
| Run 5 | 0.4562 |
| Run 6 | 0.5117 |
| Run 7 | 0.4632 |
| Run 8 | 0.4708 |
| Run 9 | 0.4698 |
| Run 10 | 0.4926 |
| Average | 0.4812 |
####2.2.2 Evaluation over all labeled valid topics
| Run | Accuracy |
|---|---|
| Run 1 | 0.5209 |
| Run 2 | 0.4631 |
| Run 3 | 0.4783 |
| Run 4 | 0.4816 |
| Run 5 | 0.4501 |
| Run 6 | 0.5055 |
| Run 7 | 0.453 |
| Run 8 | 0.4693 |
| Run 9 | 0.4763 |
| Run 10 | 0.481 |
| Average | 0.4779 |
####2.2.3 Evaluation over all labeled test topics
| Run | Accuracy |
|---|---|
| Run 1 | 0.3915 |
| Run 2 | 0.4126 |
| Run 3 | 0.4165 |
| Run 4 | 0.4749 |
| Run 5 | 0.4245 |
| Run 6 | 0.4076 |
| Run 7 | 0.3927 |
| Run 8 | 0.3655 |
| Run 9 | 0.4639 |
| Run 10 | 0.465 |
| Average | 0.4215 |