Publicação

A short survey on chatbot technology: Failure in raising the state of the art

Ver documento

Detalhes bibliográficos
Resumo:This short survey aimed initially to explore the existing state of the art for the application of chatbot on fighting (and not on spreading) of fake-news. It was then realized that there is not common to use chatbots with this “virtuous” purpose. Therefore, after two surveys and a meta-analysis, the topic had to be withdrawn since there were no survey results to discuss besides the absence of results. The survey result raised then a need to realize how chatbots are being currently used, designed and their primary sources. The result was once again confusing since, on the sample: (1) no significant concentration of usage could be found; (2) no widely adopted design strategies were identified, and (3) no significant crosscutting references to be considered as primary sources. Certainly, this can be due to a biased sample but may also be a symptom of a methodological issue on the chatbot researches. If the second possibility is proved to be right it means that chatbot research is still on a pre-paradigm stage according to Kuhn’s conception. For this paper, there were performed 4 surveys with a total sample of 50 papers mostly from the last 3 years.
Autores principais:Marcondes, Francisco Supino
Outros Autores:Almeida, J. J.; Novais, Paulo
Assunto:Chatbot research Chatbot state of the art Fake-news
Ano:2020
País:Portugal
Tipo de documento:comunicação em conferência
Tipo de acesso:acesso restrito
Instituição associada:Universidade do Minho
Idioma:inglês
Origem:RepositóriUM - Universidade do Minho
Descrição
Resumo:This short survey aimed initially to explore the existing state of the art for the application of chatbot on fighting (and not on spreading) of fake-news. It was then realized that there is not common to use chatbots with this “virtuous” purpose. Therefore, after two surveys and a meta-analysis, the topic had to be withdrawn since there were no survey results to discuss besides the absence of results. The survey result raised then a need to realize how chatbots are being currently used, designed and their primary sources. The result was once again confusing since, on the sample: (1) no significant concentration of usage could be found; (2) no widely adopted design strategies were identified, and (3) no significant crosscutting references to be considered as primary sources. Certainly, this can be due to a biased sample but may also be a symptom of a methodological issue on the chatbot researches. If the second possibility is proved to be right it means that chatbot research is still on a pre-paradigm stage according to Kuhn’s conception. For this paper, there were performed 4 surveys with a total sample of 50 papers mostly from the last 3 years.