How to visualise a conversation: case-based reasoning approach

Nikolay Burlutskiy, Miltiadis Petridis, Andrew Fish, Nour Ali

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBNpeer-review


At present, the complexity and scale of modern digital conversations between people is at its highest level but there is a gap in how to represent these conversations to a user. As a result, it is often hard for a user to understand the flow of a conversation and make an informed decision over it. However, an aesthetic and efficient visualisation can mitigate this drawback of data representation. In this paper, a case-based approach was proposed for choosing an appropriate visualisation for user’s conversations. A case was formulated as a visualisation of a conversation which a user decided to use for his analysis of the conversation. When a user decides to visualise a new conversation, the most similar visualisation type from previous users’ experiences is selected for the visualisation of the new conversation. In this paper, the cases of visualisations of conversations from the IBM Many Eyes platform were collected and a case-based reasoning approach for choosing a visualisation of user’s conversation was designed. Finally, the work of the proposed approach was tested on a sample email conversation, and then four participants evaluated the appropriateness of the chosen visualisation types in comparison with other eight possible visualisations for the email conversation.
Original languageEnglish
Title of host publication19th UK CBR Workshop, 2014
Place of PublicationCambridge
Number of pages12
Publication statusPublished - 9 Jan 2014
Event19th UK CBR Workshop, 2014 - Cambridge, 9 December 2014
Duration: 9 Jan 2014 → …

Publication series

NameCEUR Workshop Proceedings


Workshop19th UK CBR Workshop, 2014
Period9/01/14 → …


  • Case-Based Reasoning
  • Conversation Visualisation
  • Many Eyes


Dive into the research topics of 'How to visualise a conversation: case-based reasoning approach'. Together they form a unique fingerprint.

Cite this