Design and Evaluation of Euler Diagram and Treemap for Social Network Visualisation
Download as PPTX, PDF3 likes546 views
This document describes a study that evaluated different visualizations for social networks: network diagrams, Euler diagrams, and treemaps. 20 participants compared (1) Euler diagrams and treemaps representing friendship relationships, (2) representing interests, and (3) combined diagrams. Euler diagrams performed better based on ratings and error rates. For interests, Euler diagrams augmented with networks performed best. The researchers aim to address scalability, redesign treemaps, and fuse styles for more efficient social network visualization.
Design and Evaluation of Euler Diagram and Treemap for Social Network Visualisation
- 1. Mithileysh Sathiyanarayanan and Nikolay Burlutskiy University of Brighton, UK {M.Sathiyanarayanan, N.Burlutskiy}@brighton.ac.uk COMSNETS 2015 Bangalore, India Design and Evaluation of Euler Diagram and Treemap for Social Network Visualisation
- 2. Purpose of Social Network Visualisation Identifying the actors of a network Understanding relationships in a network Aggregating the information on a social network For example: network diagram (graph)
- 3. Motivation Since these days people communicate everyday via social networks such as Facebook, Twitter, LinkedIn, Instagram etc. There is a big need of visualising tools for the data analysts. However, the complexity and scale of digital communication motivates people to find alternative ways to represent social networks in visual form which are more efficient for specific reasoning tasks.
- 4. What is the challenge? In this research, visualisation of a social network using a network diagram is challenged by visualisation using a treemap and Euler diagram. The main motivations for this challenge is to understand when the two visualisations are useful and appropriate for social network visualisation and what kind of diagram combination would enhance the visualisation of social networks.
- 5. Network diagram is challenged by visualisation using a treemap and Euler diagram. Diagrams Network Diagram Euler Diagram Treemap Diagram
- 6. Addressing the Problem First, Euler diagram and treemap visualisations were designed for a Twitter network and then compared by twenty participants in an evaluation. Then two combined diagrams, an Euler diagram with network diagram, and a treemap with network diagram were designed and evaluated by the participants. Finally, the results for the evaluation of a Twitter network visualisation are demonstrated and discussed.
- 7. Design of the Study Diagrams were drawn using the software Inkscape. We considered three different sizes of curves/rectangles: small, medium and large. 20 participants were randomly selected from the University of Brighton for the study. Fifteen of the participants were male, five female, and their age varied from 21 to 29 years with 23.5 years as the mean value for their age. Next, all participants were asked questions based on the visualisations. Diagrams drawn were randomised for the participants.
- 8. Goal of the Designed Evaluation GOAL 1: To evaluate the users preference of social network visualisation from two different visualisations, namely Euler diagram and treemap. Euler Diagram Treemap Diagram
- 9. Goal of the Designed Evaluation GOAL 2: To evaluate the choice between an Euler diagram and treemap for conveying information about people’s interests. Euler Diagram Treemap Diagram
- 10. Goal of the Designed Evaluation GOAL 3: To evaluate the choice between an Euler diagram and treemap augmented by network diagrams for conveying information about people’s interests. Euler Diagram with network diagram Treemap Diagram with network diagram
- 11. Section A Comparing Euler Diagram and Treemap Representing Friendship Relationships between People How many people are in the network? How many friendship connections exist in the network? Which person has the most tweets? How many connections Adam has? Are Steve and Adam connected?
- 12. Section A: diagrams Euler Diagram Treemap Diagram Level of complexity – 6 actors. We also had 4 and 8 actors for the study.
- 13. Section B Comparing Euler Diagram and Treemap Representing People’s Interests What are the common interests between John and Luke? What are the common interests between Robin and Luke? What are the common interests between John and Steve?
- 14. Section B: diagrams Euler Diagram Treemap Diagram Level of complexity – 6 actors. We also had 4 and 8 actors for the study.
- 15. Section C Comparing two Combined Diagrams Representing People’s Interests Luke follows Tom Cruise and how many people Tom Cruise follows? Who are the common people Adam follows with Luke? How many people Luke follows which does not include common with Robin?
- 16. Section C: diagrams Euler Diagram Treemap Diagram Level of complexity – 6 actors. We also had 4 and 8 actors for the study.
- 17. Results
- 18. Discussion Results of GOAL 1: evaluation comparing Euler diagram and treemap representing friendship relationship between people, the results showed that Euler diagram performed better than treemap Euler Diagram Treemap Diagram Mean Rating (out of 5) 4.6 4.0 Error rate 4% 17%
- 19. Discussion Results of GOAL 2: evaluation comparing Euler diagram and treemap representing peoples interest, the results showed that again Euler diagram performed better than treemap. Euler Diagram Treemap Diagram Mean Rating (out of 5) 4.5 3.1 Error rate 13% 37%
- 20. Discussion Results of GOAL 3: evaluation comparing two combined diagrams representing peoples interest. The results showed that Euler diagram augmented by a network diagram performed better than a treemap augmented by a network diagram. Euler Diagram with network diagram Treemap Diagram with network diagram Mean Rating (out of 5) 4.6 3.1 Error rate 13% 35%
- 21. Future Works Other alternative ways of social network representation based on Euler diagrams and treemaps are considered for addressing the scalability issue. Redesign the augmented treemap with changed semantics and including interaction in the visualisation. Fusion of different visualisation styles is the research goal for the future since it can spark efficiency of social network visualisation. In the next study, we will use crowdsourcing for evaluation of the visualisations since the results obtained with 20 participants of this study must be verified for larger groups of participants.