Word Embedding Vectors

This paper presents an approach for estimat- ing the question-question similarity of an En- glish dataset specified in Shared Task 3, sub- task B of SemEval-2016. Given a new ques- …...

AbstractThis paper describes a system developed for a shared sentiment analysis task and its subtasks organized by SemEval-2017. A key feature of our system is the embedded ability to detect …...

AbstractThis article establishes that, unlike the legacy tf*idf representation, recent natural language representations (word embedding vectors) tend to exhibit a so-called concentration of measure phenomenon, in the sense that, as …...