Usability and utility of Small Worlds

1
Usability and utility of Small Worlds
2
The HCI problems
Results Visualization Trust ?
Trust Problem 1 Do users trust the
recommendation results shown in the
visualization? Trust Problem 2 Do users trust
the recommendation results inferred from
users interaction with the visualization?
3
The HCI problems
Simple
Understandable
Usable
Intuitive
4
Customer Us
?
?
object List something Things Stuff
VS
?
?
?
5
Methods
10 students (or more if there are
incentives) Recruitment Experiment Analysis
6

• These questions they pick a number 1-5
• How many of the top X do you like/are quality
  recommendations?
• How many of the top X people do you trust/are
  quality recommendations?
•  
• These questions they use the lichart scale. They
  do these questions once BEFORE interacting with
  the system, then again AFTER interacting with the
  system
• For the top recommended person, to what degree do
  you trust them/are they are good recommender?
• For the 5th top recommended person, to what
  degree do you trust them/are they are good
  recommender?
• I was able to understand the purpose of each of
  the five columns
• I understand the meaning of the positions of the
  nodes
• I understand the meaning of the size of the nodes
• I found the system easy to understand
• Things to try
• remove an item from your liked list. Does the
  system now recommend that item?
•  
• Tasks they need to do
• find 2 friends and increase your trust in them
  (to increase their opinion weight in your
  recommendation move 1 item from recommended to
  liked)
• When they are all done ask (for ranks, use the
  1-5 scale)

7
Task Person
Week 3 Presentation All
Week 4 Schedule room / formalize questionnaire Fang / nicholas
Week 5 Recruitment Nicholas / naiyu
Week 6 Recruitment Nicholas / naiyu
Week 7 Experiment/ analyze Fang and naiyu / all
Week 8 Presentation All
Week 9 Write report All
Week 10 Write report All