Nov 2nd 2023
In properly weighted opinion polls, support for the Palestinians is much less marked
keyboards. “At first I was angry at Hamas and Palestine for the attacks” wrote one user, “but now after seeing more of what’s going on I cannot support such a regime in Israel. #FreePalestine”.
Do such views reflect overall opinion? At our request dmr, an ai-technology firm, collected 1m posts from Instagram, Twitter and YouTube from October 7th to 23rd. All contained hashtags from a list with similar numbers of pro-Israeli and pro-Palestinian terms in English, or replied to such posts. dmr then built a machine-learning model to classify posts as backing one side, the other or neither. It was trained on content in English, but also processed posts in any language that included English hashtags.
One cause of this gap is age. Social-media users skew young, and such people are unusually pro-Palestinian in their views. Moreover, dmr’s sample did not include Facebook, which may be the most pro-Israel platform owing to its older users. In polls of people in Denmark, France, Spain and Sweden, Israel drew more total sympathy, but young participants’ opinions matched social-media ratios. Yet in America and Britain, social-media views are even more pro-Palestinian than those of young poll respondents. Israel’s backers show rather less zeal for online combat.
'Economist' 카테고리의 다른 글
| The Shibor shockChina’s central bank allows a cash crunch to worsen (0) | 2023.11.15 |
|---|---|
| Why American manufacturing is becoming less efficient (0) | 2023.11.13 |
| Markets think interest rates could stay high for a decade or more (0) | 2023.11.03 |
| Are America’s CEOs overpaid? (0) | 2023.10.30 |
| 2023 looks set to be a watershed year for the climate (0) | 2023.10.18 |