RankSQL: supporting ranking queries in relational database management systems
Chengkai Li, Mohamed Abdelrahman Soliman, Kevin Chen–Chuan Chang, Ihab F. Ilyas · 2005
Introduction Ranking queries (or top-k queries) are dominant in many emerging applications, e.g., similarity queries in multimedia databases, searching Web databases, middleware, and data mining. The increasing importance of top-k queries warrants an efficient support of ranking in the relational database management system (RDBMS) and has recently gained the attention of the research community. Top-k queries aim at providing only the top k query results, according to a user-specified ranking function, which in many cases is an aggregate of multiple criteria. The following is an example top-k query. Example 1: This is a trip planning query in PostgreSQL syntax. The user wants to stay in a hotel, have lunch in an Italian restaurant (condition c 1 : r.cuisine=Italian), and walk to a museum after lunch; the hotel and the restaurant should cost less than $100 (c 2 : h.price+r.price 100); the museum and the restaurant should be in the same city area (c 3 : r.area=m.area). Results qualifyi