Skip to main content

Show HN: Elasticsearch Learning to Rank Plugin https://ift.tt/2Z7adyT

Show HN: Elasticsearch Learning to Rank Plugin For those who don't know, Learning to Rank, is a means of using a machine learning model to optimize relevance of search results. The Elasticsearch Learning to Rank plugin creates the infrastructure for feature storage (aka templated Elastic queries), feature logging, and then uploading models trained offline for ranking with those features. This plugin is a few years old, but as one of the core maintainers, I'm eager to open the pool of contributors to this codebase used a lot by the Elastic community. I realize Elasticsearch plugins are a dark art. But if your team uses Elasticsearch for search, is considering or using Learning to Rank, we'd love to have you in the community of maintainers that includes Wikimedia, Yelp, and other big Elasticsearch deployments. The Code: https://ift.tt/2kDSRDO The Docs: https://ift.tt/2PEBdTc December 19, 2019 at 08:00AM

Comments

Popular posts from this blog

Launch HN: Simmer (YC W19) – Reviews for Delivery Dishes https://ift.tt/2Y4sD67

Launch HN: Simmer (YC W19) – Reviews for Delivery Dishes Hi HackerNews community! We’re Vaibhav and Richard, founders of Simmer ( https://usesimmer.com ). Simmer aggregates dishes from DoorDash, Caviar, GrubHub, and basically every other delivery platform out there, and tells users the best dishes across the board. We do this by providing dish-level reviews. Users find highly reviewed dishes on Simmer, choose the delivery platform of their liking, and we deep link them into their chosen delivery app. We essentially lead gen to delivery. We started working on Simmer because we loved trying new restaurants, but never knew what to order. Especially, when there’s a 5 page menu. We always asked ourselves, “Why aren’t there ratings for individual dishes?” We launched our app with this restaurant-centric use case (reviews for every dish at every restaurant), but when we rolled out delivery integrations, we realized that users resonated most with the delivery angle. They found it particularly ...