We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computers and Society

Title: Pachinko Prediction: A Bayesian method for event prediction from social media data

Abstract: The combination of large open data sources with machine learning approaches presents a potentially powerful way to predict events such as protest or social unrest. However, accounting for uncertainty in such models, particularly when using diverse, unstructured datasets such as social media, is essential to guarantee the appropriate use of such methods. Here we develop a Bayesian method for predicting social unrest events in Australia using social media data. This method uses machine learning methods to classify individual postings to social media as being relevant, and an empirical Bayesian approach to calculate posterior event probabilities. We use the method to predict events in Australian cities over a period in 2017/18.
Subjects: Computers and Society (cs.CY); Social and Information Networks (cs.SI); Applications (stat.AP); Machine Learning (stat.ML)
Cite as: arXiv:1809.08427 [cs.CY]
  (or arXiv:1809.08427v1 [cs.CY] for this version)

Submission history

From: Lewis Mitchell [view email]
[v1] Sat, 22 Sep 2018 11:40:22 GMT (401kb,D)

Link back to: arXiv, form interface, contact.