SIGMOD Edgar F. Codd Innovations Award
for innovative and highly significant contributions of enduring value to the development, understanding, or use of database systems and databases.
Dr. Umeshwar Dayal is the recipient of the 2010 SIGMOD Edgar F. Codd Innovation Award for a succession of pioneering, influential contributions in distributed heterogeneous databases, high-performance active databases, generalized transitive closure, transaction models for long-running activities, and business process discovery, among other topics.
Details: Umeshwar Dayal has a track record of 30 years of research accomplishments in data management and has made a succession of fundamental contributions to the field. His research in the mid-1980's on Multibase, the world's first large-scale heterogeneous distributed database system, paved the way for research over the past two decades at universities and industrial research labs (HP Labs, IBM Research, and elsewhere) in data integration and federated databases. PROBE, one of the first extensible database management systems, made significant contributions to the field in several areas, including object and semantic data modeling, spatial, temporal and recursive query processing, and system architecture. HiPAC was one of the leading active database systems of its time, and the only one to focus on the needs of real-time applications, resulting in an innovative transaction model. The event-condition-action (ECA) rule model introduced in HiPAC has now been widely adopted in reactive computing systems, complex-event-processing systems, and distributed middleware.
Dayal also has significant results in query-processing research, with particularly strong contributions to the processing of multi-database queries, spatial queries, and recursive queries. The approach described in his VLDB 1987 paper on unnesting SQL queries was later adopted in at least five commercial products. In addition, Dayal performed pioneering work in long-duration transactions, business-process management, and database design. In particular, he pioneered the field of business-process intelligence, which combines data warehousing, data mining, analytics and optimization techniques to monitor, control, analyze, and optimize business processes.
Over 160 research papers and over 25 patents testify to Dayal's innovation and productivity. In 2001, he received the prestigious 10-year best paper award from VLDB for his paper on a transactional model for long-running activities. He is an HP Fellow, recognized for career contributions that "caused substantial change" in the state of the art while also improving HP products. In addition to his many innovative technical contributions, he has a distinguished record of service to our community, as a member of the editorial board of major journals (including ACM TODS and VLDB), chairing conferences (including SIGMOD, VLDB and ICDE), serving on boards and steering committees (VLDB, IEEE TC on Electronic Commerce, SIAM International Conference on Data Mining among them), and mentoring junior colleagues and young researchers.
SIGMOD Contributions Award
for outstanding and sustained services to and promotion of the database field through activities such as education, conference organization, journals, standards, and research funding.
Dr. David Lomet is the recipient of the 2010 SIGMOD Contribution Award for his outstanding leadership as the Editor-In-Chief of the IEEE Data Engineering Bulletin, a key forum for dissemination of emerging ideas in academia and industry. Lomet has been a key figure in our field, holding many additional leadership roles and demonstrating in each his dedication to service and to our community.
Details: By awarding David Lomet the ACM Contributions Award, we recognize his outstanding contributions to our community in leading the IEEE Data Engineering Bulletin for nearly 20 years and thereby creating a collection of timely articles of great value. He has almost single-handedly driven the IEEE Data Engineering Bulletin, providing our community with a constant stream of special issues, assembled by world-class invited guest editors. This service has been a wonderful benefit to the field, as each issue has provided a "root node" into key projects, both academic and industrial, and into the research literature related to the topic of the issue. These articles and issues have thus provided a "fast path" to see what's happening in an area as well as a way to make sure industrial highlights are brought to the attention of academics and vice versa. Not only has Lomet run the Bulletin, but he also negotiated with IEEE Computer Society and authors to make issues of the Bulletin available on CDROM via the SIGMOD DiSC, and later, to digitize the entire set of issues from 1977 on so that they can be available online to all.
Lomet has made significant contributions to our field through service in other roles as well. He was a Co-PC Chair of VLDB and he is currently on the VLDB Board of Trustees. He has been both a Co-PC Chair and a Co-General Chair of IEEE Data Engineering Conferences, and served as a member of the Steering Committee of the IEEE Technical Committee on Data Engineering from 2004-2009. He sets a high standard of service for our community.
SIGMOD Test-of-Time Award
for the paper from the 1999 SIGMOD Conference that has had the most impact (research, products, methodology) over the intervening decade.
NiagaraCQ: A Scalable Continuous Query System for Internet Databases
Jianjun Chen, David J. DeWitt (University of Wisconsin, now Microsoft), Feng Tian (University of Wisconsin, now VMWare), Yuan Wang (University of Wisconsin, now Microsoft)
This paper from the SIGMOD 2000 Conference bridged from the world of continuous, or standing, queries against a changing stored database, to stream processing systems. NiagaraCQ was a pioneering system, the first to address the problem of the millions of overlapping queries that would need to be supported in a truly internet-scale system. It used relational-style operators to optimize a given set of continuous queries. Similar frameworks appeared in subsequent studies of stream databases, sensor databases, information delivery systems, and complex-event-processing (CEP) systems. The idea of dynamic optimization of continuous queries leveraging database operators (including dynamic query grouping and split) became a baseline for modern streaming data platforms. In summary, this paper helped open the new field of high-performance systems for continuous query processing, and was a strong force in shaping the following generations of stream processing systems.
Original abstract: Continuous queries are persistent queries that allow users to receive new results when they become available. While continuous query systems can transform a passive web into an active environment, they need to be able to support millions of queries due to the scale of the Internet. No existing systems have achieved this level of scalability. NiagaraCQ addresses this problem by grouping continuous queries based on the observation that many web queries share similar structures. Grouped queries can share the common computation, tend to fit in memory and can reduce the I/O cost significantly. Furthermore, grouping on selection predicates can eliminate a large number of unnecessary query invocations. Our grouping technique is distinguished from previous group optimization approaches in the following ways. First, we use an incremental group optimization strategy with dynamic re-grouping. New queries are added to existing query groups, without having to regroup already installed queries. Second, we use a query-split scheme that requires minimal changes to a general-purpose query engine. Third, NiagaraCQ groups both change-based and timer-based queries in a uniform way. To insure that NiagaraCQ is scalable, we have also employed other techniques including incremental evaluation of continuous queries, use of both pull and push models for detecting heterogeneous data source changes, and memory caching. This paper presents the design of NiagaraCQ system and gives some experimental results on the system's performance and scalability.
SIGMOD Best Paper Award
FAST: Fast Architecture Sensitive Tree Search on Modern CPUs and GPUs Changkyu Kim, Jatin Chhugani, Nadathur Satish (Intel), Eric Sedlar (Oracle), Anthony Nguyen (Intel), Tim Kaldewey (Oracle), Victor Lee (Intel), Scott Brandt (University of California, Santa Cruz ), Pradeep Dubey (Intel)
Citation by the Awards Committee: This paper presents FAST, a layout for an in-memory binary tree index that is well-suited for state-of-the-art CPU and GPU architectures. The layout and associated search methods take advantage of SIMD instructions and thread-level parallelism (TLP). FAST also accounts for cache-line sizes and hides cache-miss and TLB-miss latency by processing many outstanding queries simultaneously (with software pipelining and TLP). The paper shows that with all these optimizations, search on GPU is compute bound and search on a CPU is bandwidth bound. To optimize the latter further, the paper presents a key-compression scheme, which also takes advantage of SIMD instructions, to alleviate bandwidth limits and handle larger keys. Experiments show how CPU and GPU perform on trees with different sizes, how many concurrent queries are needed to achieve their peak throughput, and how compression can improve search performance. This paper is an excellent research contribution that provides an end-to-end system design and associated algorithms and techniques to develop a complete solution that leverages the underlying hardware architecture. Given the modular structure of the overall design, the solution can easily be adapted to future architectures.