If statistics/machine learning is so powerful, why can’t we predict suicidal cases in India – primarily in farmers?


If statistics/machine learning is so powerful, why can’t we predict suicidal cases in India – primarily in farmers?

I don’t think this is about the limitations of technology, or technologists.  This is about political will.  It’s not even about preventing suicides; it’s about preventing the causes of suicides.  So what this is really about is the political will to recognize and eliminate the factors contributing to increased suicides.

In terms of technology, it could be approached objectively or subjectively.  Objectively, the epidemiology can, and probably is, being tracked via geographic information systems (GIS), in order to identify not only likely hotspots, but also surrounding dynamics.  Subjectively, one of the issues in rural India is of course illiteracy; therefore, instead of usual textual approaches, innovative speech based approaches would need to be pioneered.

References:

  • A controlled trial using natural language processing to examine the language of suicidal adolescents in the emergency department (2015)
  • Detecting suicidal ideation in Chinese microblogs with psychological lexicons (2014)
  • Mood disorder patients’ language features on their microblogs (2014)
  • Suicide Prevention and New Technologies (2013)
  • Suicide note classification using natural language processing: A content analysis (2010)
  • Using natural language processing to classify suicide notes (2008)