As you read the paper, consider the following questions:
Question 1: What are the reasons for adding native XML support to relational DBMSs? Why not simply shred XML into relational tables or store XML documents as LOBs (Large Objects)?
Question 2: What features must be added to a relational DBMS to provide native support for XML data and integration with relational data?
Question 3: What are the indexing requirements for XML data? What are the similarities and differences with indexing relational data?