Well that’s easy. It just depends on what the population believes. Take a sample of the population. You don’t want extremist outsider opinions, so make sure to only include ID’d citizens. Better yet, just poll your bureaucrats and soldiers so you know you get a good sample. Then use an ML labeling technique to sort them by sentiment. Assign the sentiments on a scale from “the government needs more corporate advisors” to “the government is doing awesome.” We want to hone in on the signal so take the middle 1σ of the bell curve. Save those individual’s comments to an excel file. Send anyone who posts anything that’s not in the database to jail for misinformation.
Well that’s easy. It just depends on what the population believes. Take a sample of the population. You don’t want extremist outsider opinions, so make sure to only include ID’d citizens. Better yet, just poll your bureaucrats and soldiers so you know you get a good sample. Then use an ML labeling technique to sort them by sentiment. Assign the sentiments on a scale from “the government needs more corporate advisors” to “the government is doing awesome.” We want to hone in on the signal so take the middle 1σ of the bell curve. Save those individual’s comments to an excel file. Send anyone who posts anything that’s not in the database to jail for misinformation.
The data scientist in me is happy while the citizen is very sad.