Social networks of lexical innovation

Investigating the social dynamics of diffusion of neologisms on Twitter

Quirin Würschinger
LMU Munich
[email protected]

This repository contains the code used for the paper Social networks of lexical innovation, submitted to Frontiers in Artificial Intelligence.

Results

The results presented in the paper can be reproduced from the code in this notebook.

Processing

The results are based on the processing of all tweets using src/processing.R. This script performs the full analysis for each neologism in the sample, making use of several other scripts contained in src/ to ...

• load the tweets: load_data()
• post-process the tweets: postproc()
• perform the frequency-based analyses
  • determine the age of neologisms: get_age()
  • calculate their volatility using the coefficient of variation: get_coef_var()
  • bin tweets in monthly intervals: bin_uses()
  • plot their usage frequency: plt_uses()
• perform the social network analysis
  • determine diffusion offset: get_diff_start()
  • determine subset boundaries: get_slice_time()
  • for each subset
    • extract edges: extract_edges()
    • extract nodes:
      • sources: get_sources()
      • targets: get_targets()
      • list of all nodes: get_nodes()
    • create the network: create_net()
    • add node information: add_node_info()
    • extract network metrics: get_net_metrics()
    • export edges: net_to_gephi()