We can talk in class?!


Lab Reports, Networks, Visualization

Introduction

This dataset was found using casos.cs.cmu.edu, this data stood out to me because this was research done in 1992 where social media wasn’t as big as it is today. These data were collected by Steve Borgatti, Russ Bernard, Bert Pelto and Gery Ryan at the 1992 NSF Summer Institute on Research Methods in Cultural Anthropology. This was a 3 week course including 14 carefully selected students. 

The data was collected by placing each participant’s name on a name tag and asking each respondent to sort the cards in order of how much interaction they had with that person since the beginning of the course.

Inspiration

When beginning this project I wasn’t sure what type of network I was going to display, I found it quite interesting to see the vast amount of data that was collected. A unique network I came across was that of a chimpanzee and its family. Another I saw that caught my eye was from the hit animated show “The Simpsons”.

Simpson’s Social Network

Materials

Gephi –  is the leading visualization and exploration software for all kinds of graphs and networks. This software was used to organize my data, alter the size of my nodes, visually describe 

Microsoft Excel – is the industry leading spreadsheet program, a powerful data visualization and analysis tool. This program was used to organize and structure my data by Source, Type, Weight, and Target. 

Methodology

First, I downloaded the data, it was later opened in excel where I then organized the data. Each column needed to be labeled correctly before the spreadsheet was imported into Gephi. Columns were labeled Source, Type, Target, and Weight. Once that was completed I uploaded my excel sheet into Gephi. Upon uploading the data Gephi already form a visualization that was set by default the only problem was viewers will be unable to recognize anything. 

Edge Chart
Fruchterman Reingold Layout

I selected “Fruchterman Reingold” that automatically shaped my visual into a modern day see through rubber band ball. This layout made it easy for edges and nodes to be as visible as possible. 

By uploading my spreadsheet as Edge data, Gephi was able to produce my nodes table independently. My nodes data needed more altering than my edge data. I had to find a way of displaying the data to my viewers in a way that made sense and was easy to follow. I knew that my data dealt with people of different sex therefore I knew I wanted to represent my nodes with a femmenine color and masculine color. 

Color Key Palette
Nodes Chart

In order to identify my nodes I decided to add a colum labeled “sex” where I distinguished bewtween the Male and Females in my project. Each name was labeled with an “F” or “M” . To get a better understanding I changed my layout to ForceAtlas 2 and under Behavior Alternatives  I selected Dissuade Hubs/Prevent Overlap. This allowed me to view each node which represented a participant from the research and also be able to identify who was the most social and least social. 

ForceAtlas 2

Each node you see in this shot represents a participant. The sizes vary based on “weight” . The weight tells us who was more social than who. The larger the node the more social that individual was.

Least Social Male
Most Social Male

Results

Overview

Overall, you can see that blue edges are social connections between the males of the group. Purple indicates the social connections made between the male and females of the group. Whereas pink indicates the social connection between the females of this group. This lab was my favorite label, I was put in a place of being uncomfortable where I had to figure things out on my own. Moving forward I would like to do a much bigger social network of my favorite animated series Avatar the Last Airbender.