4 ms·
actually, the API was about being enough so that someone could write a nice client for your bookmarks. we did actually build a recommendation engine (fell down
by joshu 14y ago
actually, the API was about being enough so that someone could write a nice client for your bookmarks.
we did actually build a recommendation engine (fell down due to scale) and popular within tag feeds (that were quite good, but the d2 redesign broke them after I left.)
text analysis and feature extraction is very different than tagging.
- platz 14y agoAre there any examples of machine learning and tagging beingused together in an online bookmarking service?