This document discusses implementing information retrieval systems using relational database management systems (RDBMS). It describes representing structured document metadata and inverted indexes of terms and their frequencies in relations. It also summarizes approaches for supporting Boolean, proximity, and relevance ranking queries using SQL queries, user-defined functions, and workarounds for limitations of commercial RDBMSs. The key benefits of using RDBMSs for IR include recovery, performance, data migration, and access control.