Development Dataset

AbstractThis article describes the unsupervised strategy submitted by the CitiusNLP team to the SemEval 2018 Task 10, a task which consists of predict whether a word is a discriminative attribute …...

This paper accompanies the release of Opusparcus, a new paraphrase corpus for six European languages: German, English, Finnish, French, Russian, and Swedish. The corpus consists of paraphrases, that is, pairs …...

AbstractThis paper describes our approach to the sentiment analysis of Twitter textual conversations based on deep learning. We analyze the syntax, abbreviations, and informal-writing of Twitter; and perform perfect data …...