The document discusses sensemaking, which is how people understand and reason with data to make decisions. It describes analytic provenance, which captures the context of the sensemaking process. The author's research developed tools like SAVI and SenseMap to support sensemaking. SAVI used visualizations like timelines and maps. SenseMap captured the provenance of the sensemaking process to build a map of the user's information gathering and reasoning. This helped address open problems in navigating information and sensemaking support.
Making sense of (big) data - visual analytics and provenance
1. Making sense of (big) data -
visual analytics and provenance
Dr Kai Xu
Associate Professor in Data Analytics
Middlesex University, London, UK
k.xu@mdx.ac.uk https://kaixu.me
2. Outline
• Introduction
• What is Sensemaking
• Supporting Sensemaking: VAST Challenge 2014
• What is (Analytic) Provenance
• Navigating Big Data: SenseMap
3. Outline
• Introduction
• What is Sensemaking
• Supporting Sensemaking: VAST Challenge 2014
• What is (Analytic) Provenance
• Navigating Big Data: SenseMap
9. Outline
• Introduction
• What is Sensemaking
• Supporting Sensemaking: VAST Challenge 2014
• What is (Analytic) Provenance
• Navigating Big Data: SenseMap
10. What is sensemaking?
• Making sense of data, reasoning, and make
decisions
• It is something we do everyday:
• Plan a holiday, buy a car, understand an illness, …
• Defence, policing, investment, medical diagnosis, …
• The task is usually not well defined
• Can’t run a google search or SQL query to find the
answer
• Involve many different data
• Explorative and iterative
12. Sensemaking example: what is the
best camera for about £500?
What is the best
camera for £500?
Pixel number
Sensor size
Image quality
chromatic
aberration?!
Noise
reduction
What does
experts say?
Online reviews
What does my
friend say?
Smart phone
Compact
Full frame?
Micro 4/3?
Sony RX100
Nikon D750Samsung
Galaxy S7
What are the
price?
How do I
compare? Panasonic
LX100
Form factor
Models
14. Outline
• Introduction
• What is Sensemaking
• Supporting Sensemaking: VAST Challenge 2014
• What is (Analytic) Provenance
• Navigating Big Data: SenseMap
15. VAST Challenge 2014
• Visual Analytics Science & Technology (VAST) Challenge
• Provide dataset and analysis tasks
• Entry: visual analytics systems
• VAST Challenge 2014 – Mini Challenge 3
• Data: tweets
• Task: solve a kidnap case
Making sense of data, reasoning, and make decisions
The task is usually not well defined
Can’t run a google search or SQL query to find the answer
× Involve many different data
Explorative and iterative
21. SAVI – Message Timeline
Nguyen, P. H., Xu, K., Walker, R., and Wong, B. W. TimeSets: Timeline
visualization with set relations. Information Visualization (2016)
30. Discussions
• Sensemaking is not just one visualisation
• Timeline, entity list, map, visual narrative, etc.
• Many aspects of the underlying data
• The visual analytics tool needs to support different
types of sensemaking tasks
• Help identify starting point
• Capture/record findings
• Connection between different views/data items
• Reasoning and construct narrative
31. Outline
• Introduction
• What is Sensemaking
• Supporting Sensemaking: VAST Challenge 2014
• What is (Analytic) Provenance
• Navigating Big Data: SenseMap
32. Provenance
• What is provenance?
‘A record of ownership of a work of art
or an antique, used as a guide to
authenticity or quality’
• Data Provenance:
• Data source and collection
• Data processing and analysis
• This is not that new
• Meta-data
• Context
• Why should we care?
Who?
When?
Where?
hoW?
Why?
What?
Which?
The 7 ‘W’
34. Analytic provenance
• How users make sense of data
• The data and visualisations used
• The findings
• The hypothesis and evidence
• The reasoning process and conclusions
• Example: Should we invade Iraq?
• It is always easier after the event
• The actual question: was it the right
decision to invade Iraq with the
information and time available then.
• Can be used to support Sensemaking
Who?
When?
Where?
hoW?
Why?
What?
Which?
The 7 ‘W’
35. Outline
• Introduction
• What is Sensemaking
• Supporting Sensemaking: VAST Challenge 2014
• What is (Analytic) Provenance
• Navigating Big Data: SenseMap
36. Sensemaking example: what is the
best camera for about £500?
What is the best
camera for £500?
Pixel number
Sensor size
Image quality
chromatic
aberration?!
Noise
reduction
What does
experts say?
Online reviews
What does my
friend say?
Smart phone
Compact
Full frame?
Micro 4/3?
Sony RX100
Nikon D750Samsung
Galaxy S7
What are the
price?
How do I
compare? Panasonic
LX100
Form factor
Models
37. This is usually what it looks like
after one hour
• What is relevant and what is not?
• Where are information about image quality?
• How to compare the models?
• Where did I left off two days ago?
• How do I explain to my wife?
38. SenseMap – a map of analytic
provenance
• What can you do when you don’t have a map to
start with
• Track where you have been
• Build the map as you go
39. SenseMap – A map of sensemaking
Browser enhancement
History
Map
Knowledge
Map
40. Browser View
• Sensemaking support
• highlight, take note
• set representative
image in History Map
• Provenance capture
• action type, timing,
context, relationship
44. Discussion
• SenseMap to support
browser-based online
sensemaking
• Capture and visualise
the analytic provenance
• Support information
collection and curation
• Early stage of the
sensemaking process
45. Summary
• Sensemaking is how people understand, reason,
and make decisions with data
• Analytic provenance captures the rich context of
this process: seven ‘W’s
• There are still many open problems for
sensemaking support
• Data visualisation tools based on analytic
provenance can potentially provide a solution
• SAVI (TimeSets) + SenseMap
• SenseMap2: navigating information place
46. • SenseMap2 is available as a Chrome extension
• Just search for ‘SenseMap’ in chrome web store.
• Looking for collaboration
• Research: data visualisation in general.
• Development: new tools or existing ones.
• Others: PhD, joint funding proposal, etc.
• Contact
• k.xu@mdx.ac.uk
• https://kaixu.me