A Chatbot as a Novel Corpus Visualization Tool


Bayan Abu Shawar, Eric Atwell

School of Computing, University of Leeds, UK




The classical way to overview a large data-set is to use a visualization process, which maps the data from numerical or textual form to a visual representation that our mind can easily interpret, such as: using graphical diagrams, charts, and geometric representations. In this paper we introduce a new idea to visualize a dialogue corpus using a chatbot interface tool. We developed a java program to convert a readable text (corpus) to AIML format to retrain ALICE. We use specific domains of the BNC spoken files to retrain ALICE, and visualise the data contents of these domains via chatting. Effectiveness of this visualization of the corpus using a chatbot depends on the chatting time and the size of the corpus. Our main conclusion is that it is possible to use the chatbot tool as a visualization process of a dialogue corpus, and that we can use different training corpora to build different chatbot personalities.


Corpus, Machine Learning, Visualization, Chat, Conversation, Illustration

Language(s) English; also: Afrikaans, Arabic, French
Full Paper