Hacker News Reader

It's present for me as a % match to a restaurant I'm a regular at back home. Signals I have seen in this are menu item similarity, photo count by similar reviewers, and more basic metadata.

It didn't really translate well to visiting Japan, but does great when in a new metro and not having any personal recommendations to go off.

View on HN

Good to know. I remember the Tokyo team working on menu item extraction back then, hopefully they managed to make it work. I don't believe we were looking at photo count at all - probably user submitted photos weren't a feature on Maps at the time.

The cross city problem is one of the things I've looked at a lot (I actually wrote a design doc for it at a different company). Even with really good coverage in New York/San Francisco/London/Tokyo due to our teams being based there, it was hard to get significant personalized ranking improvement. Maybe it's better now that they have a lot more data.

Reply on news.ycombinator.com