1. Collaboration Spotting: Visualisation of LHCb process data
"The analysis of large graphs plays a prominent role in various fields of research and is relevant in many important application areas. Effective visual analysis of graphs requires appropriate visual presentations in combination with respective user interaction facilities and algorithmic graph analysis methods." [Landesberger]. 
Welcome to the first implementation of the visualisation tool for LHCb process data
A. Agocs, D. Dardanis, R. Forster, J.-M. Le Goff, X. Ouvrard, D. Proios, A. Trisovic (LHCb)
CERN
[Landesberger]: Visual Analysis of large graphs: State-of-the-art and future research challenges. T. Landesberger et al. Computer Graphics Forum, Wiley, 2011, 30 (6), pp

2. Background
Collaboration Spotting (CS) is a graph-based interactive visualisation tool for multi-dimensional data networks It aims at evolving towards a suite that supports visual analytics of Big Data.
CS is particularly efficient in performing visual queries on complex and large data networks
Data Networks are stored in Neo4j Graph DBs
CS intends to maximize human visual perception of the content of multi-dimensional data networks
CS has been originally developed to analyse large networks of publications and patents using Lucene-based technology searches)
This implementation of CS addresses LHCb process data
Below a summary of CS concepts and main features of the software
Collaboration Spotting

3. Big Data Analytics Cycle (Today)
Domain expert
Software developer
Cycle is managed by
Data Scientist
Software developer
Domain expert
Source: JIOX: Intelligence Tradecraft & Analysis

4. VISION Expert at the centre of the cycle
Experts have the knowledge
Data scientists have the skills
Bring analytics to experts
“Understand” results of analytics
“Instruct” computers to perform analytics according to findings
Domain expert
Data scientists to enable experts to perform analytics by themselves

5. Collaboration Spotting Framework
The project follows the proposed conceptual framework of D. Sacha et al*.
*Human-Centered Machine Learning Through Interactive Visualization: Review and Open Challenges
Dominik Sacha, Michael Sedlmair, Leishi Zhang, John Aldo Lee, Daniel Weiskopf, Stephen North, Daniel Keim
Collaboration Spotting

6. Big Data is organised in networks
Document systems with data and metadata in Database
Database tables with metadata in schema
Big Data is distributed
Networks are not always materialised due to the distributed nature of data sources
Ex: Publications and patents metadata
Big Data is strongly interconnected

7. Networks in LHCb Neo4J DB & related schema for dependency data
Label: Network dimension
Reachability graph: Graph of connected labels (Schema)
Collaboration Spotting

8. Graph visualisation features
Selecting network dimensions
Traversing network dimensions
Graphical queries
Time/Frequency evolution
Maximizing human understanding
Viewing multiple data sources
Looking for collaborations
Sorting data
Contextual visualisation & analytics
Enhancing reasoning
The collaboration Spotting Project

9. Navigation with CS eases the visual perception of the database content
EX: Vertex x86_64-slot-goc46-opt in Neo4j,  Same in CS
Collaboration Spotting

10. Sorting is particularly easy with CS
Component view sorted by size
Collaboration Spotting

11. Using the timeline Component view sorted by size 2010 2011
Collaboration Spotting

12. CS supported Graph Visual representations
Static graph with timeline window
Node-link using different layout techniques
Clique representation (available in LHCb)
Force Atlas (available in LHCb)
Circular representation
Extra node representation (hyper-graph)
Force Atlas
Collaboration Spotting

13. Setting up user visual environment
Reachability Graph
Graph of connected dimensions
Visual analysis dimensions (user selection)
Components, steps, applications, framework, etc.
Entry graph (user specified)
Visual dimension of the front graph
Processing Pass Description
Collaboration Spotting

14. LHCb reachability graph/available
Entry graph label
Reachability graph
Available dimensions for navigation
Blue edges added to support navigations from entry graph
Collaboration Spotting

15. Entry graph in LHCb
1 vertex (all PPD) and Navigation options (as defined in the navigation graph)
Collaboration Spotting

16. Components
Collaboration Spotting

17. Applications
Collaboration Spotting

18. Frameworks
Collaboration Spotting

19. Platforms
Collaboration Spotting

20. Steps
Collaboration Spotting

21. PPD(Modularity)
Collaboration Spotting

22. PPD(Data)
Blue: Real Data
Red: Monte Carlo
Collaboration Spotting

23. Visual graph navigation concepts
Highlight clusters
Hovering:
Node egocentric view
Left click:
Access to other dimensions for a node selection
Right click:
Navigation across dimensions
Right pane
Node-based interactivity
Collaboration Spotting

24. Hovering:
Highlight clusters
Collaboration Spotting

25. Left click:
Node egocentric view
Collaboration Spotting

26. Right click: Access to other dimensions for a node selection
Collaboration Spotting

27. Right pane
Navigation across dimensions
Collaboration Spotting

28. Your feedback
We are very much interested in developing the tool further and providing you with a service that is useful to your work
PLEASE give us feedback so that we can build Use cases with you.
Contact us at:
Collaboration Spotting

29. Thank you for your attention!