Here we make publicly available SentiLex sentiment lexicon for English. The creation of the dictionary is thoroughly described in:
Glavaš, Goran, Jan Šnajder, and Bojana Dalbelo Bašić. "Experiments on hybrid corpus-based sentiment lexicon acquisition." Proceedings of the Workshop on Innovative Hybrid Approaches to the Processing of Textual Data. Association for Computational Linguistics, 2012.
Here we make publicly available evaluation sets for sentiment lexicon for Croatian.
We prepared three different evaluation sets, with different levels of inter-annotator agreement. All evaluation sets were split evenly into two sets: one for model validation and parameter fitting and other for testing. For each complete set (containing words marked with "+", "-", and "0") we prepared accompanying "polar" set, containing only words marked with "+" and "-". Therefore, there are 4 files available for each level of inter-annotator agreement: validation set, test set, polar validation set, and polar test set.
All files are encoded in UTF8 and have Windows line endings.
SentiLex - all files available by TakeLab, FER are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.