Creation of a corpus of embedded words from tweets generated in Argentina

Abstract

Text processing of any kind is a task of great interest in the scientific community. One of the social networks where people frequently express themselves freely is Twitter, and therefore, it is one of the main sources for obtaining textual data. In order to perform any type of analysis, the first step is to represent the texts in a suitable way so that they can then be used by an algorithm. This paper describes the creation of a corpus of word representations obtained from Twitter using Word2Vec. Although the sets of tweets used are not massive, they are considered sufficient to take the first step in the creation of a corpus. An important contribution of this work is the training of a model that captures the idioms and colloquial expressions of Argentina, and includes emojis and hashtags within the vector space.