7 ms·
Well done! A lot of progress since last time. Have you guys considered using AI to categorise products (ie; create labels using product images), instead of usin
by LuigiElsa 3y ago
Well done! A lot of progress since last time. Have you guys considered using AI to categorise products (ie; create labels using product images), instead of using the text to match the search? I say this cause I sometimes see some irrelevant products and I can tell you guys are basing the search on text
- pencildiver 3y agoThanks! Definitely a work in progress but getting better by the day. We have run tests with image-detection to try to categorize products. We currently do search based on the name, description, price, store, and brand. The problem with image-detection is cost. Given the size of our data set, it's very costly to run 800m - 1b images through a model (i.e. most products have 4 - 5 images). We've considered only doing the first 'hero' image to start though. Open to any cost-effective ideas though. For example, if you search for "wooden chair", it would be nice to select a filter for 'category' to narrow down if I want to see "office furniture", "dining room", or "art". https://www.searchagora.com/search?query=wooden+chair&count=20&position=productCard65b82f078c659c8464aa6e57 https://www.searchagora.com/search?query=wooden+chair&count=...
- LuigiElsa 3y agoI found some things on Github you could use, I'm not a dev myself and I'm not sure how scalable these are, but have a look, maybe there's something useful. https://github.com/jhc13/taggui https://github.com/jhc13/taggui The category filtering is what I wanted to get at, I think the search would improve a lot.
- pencildiver 3y agoSuper cool, thanks! Will check it out.