Learning GUI Design Vector Space for Assisting Developers with GUI Development
The apps(from Rico dataset: http://interactionmining.org/rico) with less than 10 GUI screenshots are filtered out, as it may mean that those apps are not fully explored. There are 2618 apps left. And we take about 84.6% of apps (2,214) for training, 5.4% (142) for validation i.e., tuning the hyperparameters of our model, and the rest 10% (262) for testing. In detail, we have 17,879 intra/inter pairs for training, 1,132 intra/inter pairs for validation, and 2,178 intra/inter pairs for testing.
Download: https://drive.google.com/drive/folders/1pNXUYAVOBiRDeix2x2Pc_dMhXfr-frC5?usp=sharing()
A pilot study to manually check the assumption for the GUI design style:
| SQuAD v1.1 Leaderboard (Oct 8th 2018) | Test EM | Test F1 |
|---|---|---|
| 1st Place Ensemble - BERT | 87.4 | 93.2 |
| 2nd Place Ensemble - nlnet | 86.0 | 91.7 |
| 1st Place Single Model - BERT | 85.1 | 91.8 |
| 2nd Place Single Model - nlnet | 83.5 | 90.1 |