| • How do different iterated algorithmic bias modes affect the boundary shift? |
• Both iterated filter bias and iterated active learning bias have a significantly effect on the boundary shift, while random selection does not have a significant effect, indicating that the nature of the model, and hence which items will be judged to be relevant to the user, changes depending on the iterated algorithmic bias, with filtering bias exerting the biggest influence. |
| • How do different iterated algorithmic bias modes affect the blind spot size? |
• There is a significant increase in the class-1 blind spot size in the testing set for iterated filter bias and active learning bias. On the other hand, there is no significant difference with random selection. |
| • How do different iterated algorithmic bias modes affect inequality of prediction? |
• Filter bias leads to a significant increase in the Gini coefficient, while both active learning and random selection show a significant decrease in the Gini coefficient. |
