Current problem of anaphoric computational processing: The case of FreeLing 4.1

• Main Authors: Andrés Felipe Grajales Ramírez, Jorge Mauricio Molina Mejía
• Format: Article
• Language: English
• Published: Universidad del Valle 2019-11-01
• Series: Lenguaje
• Subjects: computational linguistics; natural language processing; text coreference; anaphora resolution; corpus linguistics.
• Online Access: http://revistalenguaje.univalle.edu.co/index.php/lenguaje/article/view/8356/11384

In this article,we present the case of automatic computational processing of the anaphora and the problem generated by such processing. To do this, we have relied on the FreeLing 4.1tagger, and its module for the treatment of textual coreference; analyzing, in this case, a corpus for teaching Spanish to foreign students. In order to establish the analysis and problems raised by this automatic analyzer, a definition and anaphoric typology of ourown has been established, which arise from various readings and theories fromtextual linguistics field. The article concludes with the statistical analysis showing the different problems generated in the recognition of these types of anaphoraand their involvement in the subsequent tagging of coreferential networks.