Mobile shopping app Mercari wants to make selling online even easier than buying, and it’s turning to two MIT professors for help.
Mercari opened an engineering office in Cambridge, Massachusetts where two renowned professors at the adjacent Massachusetts Institute of Technology (MIT) will advise it in machine learning, computer vision, and related research.
Mercari’s team in Cambridge is applying machine learning on massive amounts of data to better personalize the user experience based on their preferences, history, and interactions, according to the company.
“Opening the Mercari engineering offices in Cambridge is an important addition to our global technology centers in Palo Alto and Tokyo,” said Dr. Mok Oh, chief technology officer for Mercari U.S. “The expectations of Mercari sellers and buyers are always evolving, so it’s vital that we harness emerging technologies to continuously improve our app and services.”
Advising the team are Dr. Frédo Durand and Dr. Wojciech Matusik of MIT.
Dr. Frédo Durand is a renowned researcher and professor in computer vision and computational photography. He is a professor of Electrical Engineering and Computer Science at MIT, and a member of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL).
Dr. Wojciech Matusik is widely-recognized professor and entrepreneur in machine learning, 3D vision, and 3D fabrications. He is a professor of Electrical Engineering and Computer Science at the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT, where co-directs the Computer Graphics Group and teaches machine learning courses
SOURCE: Mercari Press Release
How about not banning people left and right for no reason at all
I am afraid this will make the system go down the hill like Ebay and etsy.
Once you try to “custom” fit any preferences (better personalize the user experience based on their preferences, history, and interactions, according to the company.) I think you add more code to slow down the system and since people’s likes change daily – you end up seeing the same stuff you would never buy again – like, oh a party hat for a friend, a kid’s baseball mitt for a nephew, a tie for dad – so now you will be show tons of the same items again and again – assuming you need one on a daily basis.
In retrospect of other bogged down systems – this may encourage people to visit their local retailers so they can actually walk up and down the aisles to see what catches their eye – instead of the computer trying to push that toilet paper twice weekly that you buy on-line or that air filter system replacement that needs to be bought – and the “cheaper dealers” are not usually in the picture at that point. The computer shows the HIGH PRICED Leaders – so customer then goes to other websites to find the less cost alternatives.
So sad that the society has come to this. All People Don’t BUY 24/7. There are cycles weekly, monthly and yearly and some months are slower, some profitable. Trying to encourage buying tailored made ends up in coding problems that don’t fit the mobile app, the PC or the customer when new and improved additions materialize. Of course I laugh when a AI tries to make sense of a unknown Brand placed in a system – it tries to autocorrect the spelling so you end up seeing a large unwanted garbage list of what a computer thinks YOU MEAN. So the new and improved just might push people to go visit local establishments after all.