This article describes the different systems we have developed to participate at the IberLEF 2019 Portuguese Named Entity Recognition and RelationExtraction Tasks (NerReIberLEF2019). Our objective is to compare rule-basedand neural-based approaches. For this purpose, we applied our systems to twospecific subtasks: Named Entity Recognition (Task 1) and General Open Infor-mation Extraction (Task 3) in Portuguese texts.
Keywords: named entity recognition, open information extraction, natural language processing