This code repository collects info and material related to the ITAmoji 2018 organized in the context of EVALITA 2018, the 6th NLP Italian evaluation campaign.
More info at: https://sites.google.com/view/itamoji/.
- ITAmoji Python Evaluation Script: evaluationScript_P.py
- Runs submitted by participating systems in the system_runs folder (there is one sub-folder for each participating team including all the runs submitted, one file per run). The format of the run submission is explained into details at: https://sites.google.com/view/itamoji/data-and-tools
It is possible to download the train and test dataset as a password-protected ZIP archives from the following links:
- Train dataset (250k tweets): https://drive.google.com/file/d/1VSN9H3vZVTIr0qOXYPmiBp6FHsvovRlb/view?usp=sharing.
- Test dataset (with gold standard labels) (25k tweets): https://drive.google.com/open?id=1kbPGdyI3fg6oSQIjx5-91t7ZXfbyxtFu.
Both ZIP archives are protected by the same password that can be requested by sending an email with subject "ITAmoji dataset password" to itamojievalita@gmail.com.
More information on the format of both datasets can be found at: https://sites.google.com/view/itamoji/data-and-tools.