Mapstr pro
The revolution in address discovery, with recommendations that match you 100%!
With over 4 million users in 90 countries, mapstr is already your indispensable ally for saving and sharing your favorite addresses. Now we're taking it a step further with Mapstr Matching, a revolutionary technology that transforms the way you discover new addresses.
.webp)
Mapstr Matching is like having a friend in every city who knows your tastes. Wherever you are in the world, we'll suggest addresses that suit you best.
Mapstr Matching is based on your saved favorite places and, thanks to Machine Learning, compares your desires, interests and habits with those of our global community. This unique process gives you access to tailor-made recommendations, perfectly aligned with your personal tastes.
.webp)
Forget the hours spent searching for the perfect location. Mapstr Matching offers you an organized selection of places to discover and organize with your favorite tags, perfect for spontaneous or carefully planned adventures. Save time for what really matters: new experiences!
Addresses already present
We reference places all over the world!
Passionate users
Benefit from personalized recommendations and contribute to an enriching collaborative experience.
Quality partners
Get inspiration from the experts!
No more notes at the bottom of your phone that you can never find again, no more screenshots and no more posts saved on Instagram.
First make sure your Mapstr application is up to date, or update it on the Apple store.
This new feature can then be accessed from your Map via the purple "Magic Wand" button in the top right-hand corner, or from the first section of the "Discover" tab.
Mapstr Matching draws on the strength of the mapstr community and the intelligence of Machine Learning technology. This feature is based on the places a user has added to their map, and compares their desires, interests and habits with all the other users in the world to give them access to tailor-made recommendations of places, adapted to their personal tastes.
It's like having a friend in every city, who knows exactly what you like and recommends the best spots!
You've probably already seen the results of Netflix or Spotify's algorithms for recommending similar films or music.
Generally speaking, these platforms use "collaborative filtering" and often even "implicit collaborative filtering" algorithms to compare users with each other based on the films and music they've watched (without an unwatched film being perceived as a negative rating, hence the implicit).
There are lots of papers and tools explaining how to do collaborative filtering, lots of off-the-shelf solutions that could be used. Except that Netflix has around 5,000 films in its catalog... Spotify has around 60 million, with lots of analyzable criteria, such as music style, tempo, band, musicians, etc. On these platforms, the "hits" are the most important. On these platforms, the "hits" are watched or listened to by a large proportion of their users.
On the other hand, mapstr already contains more than 90 million registered places, with no factual criteria that can be organized ("French restaurant" is vague...), among which the most popular are only "added" by less than 2% of users! In terms of matrix calculation, we end up with matrices filled with 0s, and some very rare 1s, in proportions that have nothing to do with those of Netflix & co... and therefore wouldn't work with all the "classic" algorithms.
This is why Mapstr Matching is unique! It's a considerable amount of work carried out by the mapstr team, based on unique data: the mix of our users' visit intentions and approved visits.
Mapstr Matching is not yet available on Android, but will be soon - stay tuned!
.png){kind=small}
