Creation of a corpus of embedded words from tweets generated in Argentina. Cuadernos de Ingeniería, v. 13, n. XIII, p. 07–24, 13 Dec.2021.