Yesterday, I tabled one of this round’s ideas. Today, I wanted to ensure that a user’s ratings persist even when the catalog updates.
I also taught search to rank. Until today, a match in an episode description either counted or it did not, nothing in between. Now stronger matches rise, the same idea the recommender is built on: order by evidence. The slowest search also got twice as fast.
What’s next?
I am going to be starting a new job so there will be less coding for TVLens until October 1. Between now and then, I will be writing about TVLens’ vision for TV show fans.
Until then, enjoy your shows, friends!