Preface on the Iberian
Languages Evaluation
Forum (IberLEF 2019)

The goal of IberLEF 2019 is to encourage the research community to organize
competitive text processing and understanding tasks with the aim of defining
new research challenges and setting new state-of-the-art results for the Natural
Language Processing community, involving at least one of the following Iberian
languages: Spanish, Portuguese, Catalan, Basque and Galician.
    IberLEF 2019 is held together with the XXXV Congreso Internacional de la
Sociedad Española para el Procesamiento del Lenguaje Natural (SEPLN 2019)
in Bilbao, Spain, on the 24th of Sep, 2019. It includes the following nine tracks:

   • eHealth Knowledge Discovery (eHealth-KD). The objective of this track
     is to encourage the development of software technologies to automatically
     extract a large variety of knowledge from eHealth documents written in
     the Spanish Language.

   • Factuality Analysis and Classification Task (FACT). Its aim is the classi-
     fication of events in Spanish according to their factuality status.

   • Humor Analysis based on Human Annotation (HAHA). It is focused on
     the classification of the humorous meaning of Spanishs texts. The track
     proposes two tasks: Humor Detection and Funniness Score Prediction.

   • Irony Detection in Spanish Variants (IroSvA). This task is fully dedicated
     to identify the presence of irony in short messages (tweets and news com-
     ments) written in three different Spanish variants: from Cuba, Mexico
     and Spain.

   • Negation in Spanish (NEGES). Negation in Spanish (NEGES). The ob-
     jective of this track is to advance the study of the detection and treatment
     of negation in Spanish.

                                        1
2

    • Named Entity Recognition and Relation Extraction for Portuguese. It
      proposes several challenges related to Named Entity Recognition and Re-
      lation Extraction in texts written in Portuguese.
    • Authorship and Aggressiveness Analysis in Twitter: case study in Mex-
      ican Spanish (MEX-A3T). The track is focused on author profiling and
      aggressive speech detection on texts written in Mexican Spanish.
    • Sentiment Analysis Task at SEPLN (TASS). The aim of this task is to
      promote research into specific Natural Language Processing techniques
      for solving problems related to the sentiment analysis of texts written in
      Spanish.
    • Medical Document Anonymization task (MEDDOCAN). The aim of this
      track is the anonymization of medical documents in Spanish, and it is
      structured into two sub-tasks: NER offset and entity type classification
      and sensitive token detection.

   The IberLEF 2019 has had a high participation, in total 179 participants
shared out the different tracks, submitting 78 papers.

    September 2019

                                                                    The editors