Methods, systems, and apparatus, including computer programs encoded on computer storage media, for automatically generating search terms and queries and selectively executing those queries based on feedback from previous queries. In one aspect, a method using artificial intelligence, natural language processing, and machine learning techniques, includes actions of receiving an input data set and first generating a set of auto-generated terms by using a predetermined and dynamic data filter, a basic utility processor, a multi-word combination generator, and a term tally and sorting processor. Additional actions include generating a primary query for every multi-word combination and generating focus queries for every primary query. Further actions include selectively executing queries in succession on a document server and database only if necessary, based on feedback from previous queries, and aggregating all query search results.