Researchers at UC San Diego got full data from the private Instagram accounts of 415 undergraduates and used machine learning to find traces of loneliness in photos, captions and activity patterns.

You may not need to ask whether someone is lonely; how they post on Instagram might tell you. Researchers obtained every photo, caption and like from the private accounts of 415 students, then used machine learning to compare that activity with the students' self-reported loneliness scores.
The detail that struck me most: students who reported feeling lonely were more likely to use a profile photo with an upward head tilt, and their captions leaned more on past-tense words like 'was' and 'had'.
The study also split loneliness into two kinds, a lack of close emotional connection and a lack of community, and the posting habits differed clearly between them. I think a platform nudging users with a caring message based on these signals could be a good thing, as long as it does not turn into another channel for targeted ads.



Curated from high-quality sources, with concise summaries and key takeaways.