Click here for the dataviz.

This is a visualization of the network of academic-philosophy-twitter, based on the list collected by @truesciphi. On the whole, this contains the 791 professional philosophers with more than a thousand followers, only a few private accounts are missing.

The philosophers are arranged by the similarity between the lists of accounts which they mention or retweet on their timeline, based on a sample of altogether 2004543 tweets. The positions are calculated using UMAP with cosine-similarity, and some postprocessing with Gephi & Datashader.

The names are scaled according to the number of their followers, but the scaling is not linear so that small accounts remain legible. [Note that this .pdf-file can be searched for names.] Accounts that are placed close together can be expected to share similar material, and I think we arrive at some meaningful clustering. Accounts that retweet each other are linked via edges, and the larger bundles of edges hint at the intensity of interaction between the corresponding clusters.

The whole thing is as of now a fun Friday-afternoon-project, so please don’t base anything important on it.