Database System Extensions for Decision Support: the AXL Approach.
Haixun Wang, Carlo Zaniolo · 2000
Research on database-centric data mining is seeking to improve the eectiveness of database systems in decision support applications. Dierent solutions are now used for dierent problems, including (i) SQL extensions for more complex OLAP queries, (ii) new datablades for special data types such as time-series, and (iii) architectural extensions to support data mining functions. Here, we proposed a unied solution for all these problems; the solution is based on User-Dened Aggregates (UDAs) expressed in an SQL-like language called AXL. In this paper, we discuss the architecture and implementation of the AXL prototype and its use and performance in expressing data mining functions and complex OLAP queries. 1 Introduction The dicult problem of extending database systems for decision support applications is of practical importance and of great research interest. Patchwork solutions were proposed in the past for dierent problems. For instance, various datablades (a.k.a. DB extenders, c...