quelich / scientist-graph

A graph representation of scientists worked together

scientist-graph

A graph implementation to demonstrate the network between scientists working together.

Proposal

This project suggests an algorithm implementing the Graph theory for correlating a network among scientists working together. Therefore, the number of papers published by scientists working together and will be considered valuable information. Thus, the project outputs the connection details between scientists.

Data

There are two types of data: Names.txt and Networks.txt

• Names.txt file maps every scientist in the line separated into two distinct columns. Each column respectively represents the properties "Name" and "Last Name" of a scientist.
• Networks.txt file maps every connection between given scientist IDs and the connection's weight, the number of publications into three columns. The first column represents the source node. The second column represents the destination node which forms a vertex to which the source vertex will be connected. The third column represents the weight between nodes.

Components

Graph

Represents the whole network, a massive collection of the edges.

Edge

Represents the connection between a single scientist and another scientist.

Node

Represents a scientist with certain properties to describe a vertex.

Results

Note: this result only shows a small part of the output. See output.txt for more of it.

• Edge: #1:Pedro :Domingos -> #2:Daniel S. :Weld, Publications:1
• Edge: #1:Pedro :Domingos -> #3:Brigham :Anderson, Publications:1
• Edge: #2:Daniel S. :Weld -> #1:Pedro :Domingos, Publications:1
• Edge: #2:Daniel S. :Weld -> #3:Brigham :Anderson, Publications:1
• Edge: #2:Daniel S. :Weld -> #40:Jia-Yu :Pan, Publications:1
• Edge: #2:Daniel S. :Weld -> #193:Peter :Antal, Publications:1
• Edge: #2:Daniel S. :Weld -> #324:Patrick :Wendel, Publications:1
• Edge: #2:Daniel S. :Weld -> #325:Igor V. :Cadez, Publications:1
• Edge: #2:Daniel S. :Weld -> #326:Scott :Gaffney, Publications:1
• Edge: #2:Daniel S. :Weld -> #327:Matt :Cutler, Publications:1
• Edge: #2:Daniel S. :Weld -> #503:D. :Gershkovich, Publications:3
• Edge: #2:Daniel S. :Weld -> #504:Ehud :Gudes, Publications:2
• Edge: #2:Daniel S. :Weld -> #625:Kamal :Nigam, Publications:1
• Edge: #2:Daniel S. :Weld -> #778:Marcel :Holsheimer, Publications:1
• Edge: #3:Brigham :Anderson -> #1:Pedro :Domingos, Publications:1
• Edge: #3:Brigham :Anderson -> #2:Daniel S. :Weld, Publications:1
• Edge: #3:Brigham :Anderson -> #2124:Thore :Graepel, Publications:1
• Edge: #3:Brigham :Anderson -> #2125:Ralf :Herbrich, Publications:1
• Edge: #4:Andrew W. :Moore -> #5:Andrew :Connolly, Publications:1
• Edge: #4:Andrew W. :Moore -> #6:Robert :Nichol, Publications:1

Conclusion

• Graph theory is favourable to describe a network where the connection between nodes are enumerated and assigned to weights.

What I learned

• Implement Graph Theory for a real world data.
• Data manipulation in C#