API and datasets to evaluate the proposed methodology at 'A Methodology to Handle Social Media Posts in Brazilian Portuguese for Text Mining Applications'
http://brs-nlp-api.appspot.com/
Text Categorization: Dataset crawled from Twitter, composed by 600 tweets from 2013, 2014 and 2015, categorized between 6 topics related to "beer" domain (manually annotated by 2 specialists).
Opinion Mining: Reviews from Google Play (provided by: F. Santos e M. Ladeira, “The Role of Text Pre-processing in Opinion Mining on a Social Media Language Dataset”, Brazilian Conference on Intelligent Systems, p. 50-54, 2014.)
https://lucene.apache.org/core/
http://developer.cybozu.co.jp/archives/oss/2010/10/language-detect.html
https://github.com/haifengl/smile