Non-contextual classification of offensive comments

Classification of  comments into different categories with contextual information. No information about the comment (source or influencer ID) is provided. 

The following categories are considered:

  • Offensive, target is a person (OFP). Offensive text targeting a specific individual.
  • Offensive, target is a group of people or collective (OFG). Offensive text targeting a group of people belonging to the same ethnic group, gender or sexual orientation, political ideology, religious belief or other common characteristic.
  • Offensive, target is different from a person or a group (OFO). Offensive text where the target does not belong to any of the previous categories, e.g., an organization, an event, a place, an issue.
  • Non-offensive, but with expletive language (NOE). A text that contains rude words, blasphemes or swearwords but without the aim of offending, and usually with a positive connotation.
  • Non-offensive (NO). Text that is neither offensive nor contains expletive language.
Publication
Flor Miriam Plaza-del-Arco, Marco Casavantes, Hugo Jair Escalante, M. Teresa Martín-Valdivia, Arturo Montejo-Ráez, Manuel Montes-y-Gómez, Horacio Jarquín-Vásquez, Luis Villaseñor-Pineda (2021) Overview of MeOffendEs at IberLEF 2021: Offensive Language Detection in Spanish Variants.Procesamiento del Lenguaje Natural, Revista nº 67, septiembre de 2021, pp. 183-194.
Language
Spanish
NLP topic
Abstract task
Dataset
Year
2021
Ranking metric
Micro F

Task results

System MicroPrecision MicroRecall MicroF1 Sort ascending
NLP-CIC 0.7679 0.7093 0.7324
UMUTeam 0.7861 0.6919 0.7301
GDUFS DM 0.7565 0.7002 0.7239
Marta Isabel 0.5781 0.5451 0.5595

If you have published a result better than those on the list, send a message to odesia-comunicacion@lsi.uned.es indicating the result and the DOI of the article, along with a copy of it if it is not published openly.