IBM Research has been a leading innovator in data management for more than three decades. In 1970, E. F. Codd of IBM Research published the landmark paper, "A Relational Model of Data for Large Shared Data Banks", which created the relational model that underlies most database systems today. Three years later, IBM Research began work on the seminal System R relational database management system (RDBMS) that provided the first implementation of the Structured Query Language (SQL), as well development of critical RDBMS technologies for query compilation, cost-based optimization, ad hoc query formulation and on-line data definition.
Over the years, key innovations from IBM Research followed, including Starburst technologies for extending RDBMSs to handle new forms of information; the Garlic framework for data federation across diverse data management systems; the QBIC algorithms for content-based querying of images and the ARIES (Algorithms for Recovery and Isolation Exploiting Semantics) family of transaction locking technologies, which underpins the database industry today.
IBM Research is exploring emerging requirements in several new areas:
- RDBMSs -- scaled to larger databases and numbers of simultaneous users based on utility-based grid computing paradigms.
- DB2 support for standards-based Web services and the IBM e-business on demand business model.
- IBM's DB2 Universal Database system using the powerful XQuery language to handle XML content.
IBM Research continues to expand its world-class data management research by exploring novel approaches for making database systems self-managing, or "autonomic", hence achieving higher performance with lower cost of administration. It is also integrating information across heterogeneous data sources, including support for real-time event processing and data streaming environments. And it is extracting knowledge from structured, unstructured, multimedia and sensor data.