Search engines can be used for a variety of high-value machine learning applications. Practical machine learning involves more than just machine learning itself. It is also necessary to be able to deploy, monitor, manage, and understand the operation of any advanced analytics application. Recent developments in machine learning allow search engines such as Lucene and Solr to be used to deploy such advanced analytics systems. This has many practical benefits because search engines based on Lucene collectively have nearly a mega-year of run-time and are thus correspondingly stable.
In this session, Joe Blue, Data Scientist, will describe how a surprisingly small amount of code can be used to deploy a recommendation engine on top of Solr. All required code is available as open source and after this talk, most implementors will have enough information to implement a production quality recommendation engine in a few days of effort. He will also have additional handouts at the talk to assist developers.