Nils Droste
2018-12-14
Droste, N., D'Amato, D., Goddard, J.J. (2018) Where communities intermingle, diversity grows - The evolution of topics in ecosystem service research. PLoS ONE 13(9), doi: 10.1371/journal.pone.0204749
We downloaded all “ecosystem service*” literature
Unsupervised machine learning: latent Dirichlet allocation of topic \( z \)
\[ P(w_i|d_k)= \sum_{z=1}^Z P(w_i|z)P(z|d_k) \]
with \( w_i \) words and \( d_k \) documents, drawing \( z \) until model converges.
we then tracked content over time / across periods
Global awareness
ES research has a consistent and large natural science base
social ES research is self-reflected and governance oriented
Based on a machine-learning review of >14k articles on ecosystem service (ES), we find
— Nils Droste (@NilsDroste) September 29, 2018
1) ES research has a consistent and large natural science base
2) social ES research is becoming more integrative and governance oriented https://t.co/cLJ1FcGQUi @Dalia_DAmato & Jess Goddard