K Nearest Neighbor Queries and KNN-Joins in Large Relational.pdfVIP

K Nearest Neighbor Queries and KNN-Joins in Large Relational Databases (Almost) for Free Bin Yao, Feifei Li, Piyush Kumar Computer Science Department, Florida State University, Tallahassee, FL, U.S.A. {yao, lifeifei, piyush}@ Abstract— Finding the k nearest neighbors (kNN) of a query no restaurants offer both Italian food and French wines) and point, or a set of query points (kNN-Join) are fundamental the kNN retrieval is not necessary at all. It is well known problems in many application domains. Many previous efforts to that when the data is stored in a relational database, for solve these problems focused on spatial databases or stand-alone systems, where changes to the database engine may be required, various types of queries using only primitive SQL operators, which may limit their application on large data sets that are the sophisticated query optimizer built inside the database stored in a relational database management system. Furthermore, engine will do an excellent job in finding a fairly good query these methods may not automatically optimize kNN queries or execution plan [4]. kNN-Joins when additional query conditions are specified. In These advantages of storing and processing data sets in a this work, we study both the kNN query and the kNN-Join in a relational database, possibly augmented with additional query relational database motivate us to study the kNN query and conditions. We search for relational algorithms that require no the kNN-Join problem in a relational database environment. changes to the database engine. The straightforward solution Our goal is to design algorithms that could be implemented