Blog Post

Network Analysis Inquiries

Although my DSC project does not explicitly involve networks, I’ve been interested in learning more about network analysis for other projects—most of which involve social media. A couple years ago I tried to analyze #pegida but ran into several issues that forced me to scrap the network analysis part of the project. With this past experience and the workshop in mind, I have a couple questions:

1. Is it possible to retrieve old tweets? I found a free tool, Twitter Archiver, to scrape for my data. However, the movement was already a month old when I started the scraping process, and I couldn’t find a way to retrieve older tweets. I felt I missed an opportunity to really understand the emergence of this movement because I wasn’t in the right digital space at the right time. Are there now tools that allow for such retrieval or has Twitter kept the lid on this? Sadly, I know I can’t rely on the Library of Congress’s Twitter project to resolve this anytime soon.

2. Is there a way to do a network analysis among hashtags rather than users? Sometimes what is most interesting is not who is being connected but what ideas are being connected. It seems helpful to map the relationships between hashtags. Playing around with NodeXL, I found different ways of changing the properties of the individual nodes, but I couldn’t figure out how to create a hashtag network. Searching for an answer I came across hashtagify.me and quickly tried it out with #notmypresident. While visually interesting, how is this network map actually produced?

80

1 comment

Matt knows this better than me, but Twitter only makes recent data available through their API. There are commercial services that have all of Twitter archived, and we have been talking about subscribing to one of these services.

Pretty interesting stuff about hashtag networks. Hastagify.me is a commercial product for marketers--I wonder if there's anything else out there?

38