In IBM Rank and Retrieve, is it possible to use your own Solr Cluster to save on expenses? If so, how? I am creating a chatbot and require Retrieve and Rank for the same. The IBM free service seems to have only a 50MB limit to the solr cluster. Is there any way to bypass this?
It is possible to use the rank portion of the Retrieve and Rank portion as a standalone learning-to-rank component. However, this requires having to generate all your own features on top of Solr. The feature generation part of the Retrieve and Rank service is only available in IBM's Solr Cluster...so it can't really be used with your own Solr Cluster.