A Process for Extracting Knowledge Base for Chatbots from Text Corpora
Aliane Loureiro Krassmann, João Marcos Flach, Anita Raquel Cestari da Silva Grando, Liane Margarida Rockenbach Tarouco, Magda Bercht · 2019
This paper presents a process for extraction of AIML chatbot knowledge bases from text corpora using the AGATA (Automatic Generation of AIML from Text Acquisition) system, developed to facilitate and reduce the time needed in that task. In order to illustrate how the process works, a Pilot Study was carried out, which consisted of converting the content of two textbooks into the AIML language and sending the files to a chatbot knowledge base, being finally assessed by experts in the domain of the content. The results show the potential of helping teachers to easily create their own chatbots according to their course topics.