BM25 is So Yesterday: Modern Techniques for Better Search Relevance in Solr
summary
Shows a practical pipeline that layers Solr 6.5 + Spark 2.x: index enrichment (NER, word2vec, topic tags), capture click signals into a clickstream sidecar, train models offline and push features back into Solr, and re-rank a coarse candidate set with learning to rank (applied to the top 100 re-rank window). Demonstrates an ALS recommender, intent classification used as a query filter, and claims a #5 finish in a relevance competition within three days using these techniques.
AI-written summary · about
chapters (7)
- 0:00Start
- 5:16BM25 (aka Okapi)
- 11:49But What About the Real World? Indexing Edition
- 15:13Real World? Query Edition
- 32:40Experimentation, Not Editorialization
- 35:06Lucidworks Architecture
- 37:30Demo Details
related talks
40mLevel Up: Everything you need to know about search, ...
43m[Machine Learning] Creating more relevant search results with Learn to Rank
43mLearning to Rank for Reddit Search: A Project Retro
46mDiscovery & Analytics with Hadoop, Mahout & Solr
37mSearch and Recommendations: 3 Sides of the Same Coin #bbuzz
23mMachine Learning for Apache Solr, Elasticsearch & Vespa #bbuzz
26mStreamlining Property Search Across Multiple Platforms
38mLearning to Rank: where search meets machine learning #bbuzz