Tempdb

Snapshot Semantics for Temporal Databases

Snapshot and sequenced semantics is widely used for evaluating queries over temporal data. Under snapshot semantics, temporal relations are seen as sequences of snapshot relations, and queries are evaluated at each snapshot. Sequence semantics extends this concept by allowing queries to access time in a controlled manner. In this work, we investigate these semantics for multiset relations and other extensions of the relational data model such as provenance, incomplete and probabilistic databases, and more.

As a first result, we have demonstrated that current approaches for snapshot semantics over interval-timestamped multiset relations are subject to two bugs regarding snapshot aggregation and bag difference. To remedy this situation, we have introduced a novel temporal data model based on K-relations that overcomes these bugs and prove it to correctly encode snapshot semantics. Furthermore, we have developed an efficient implementation of our model as a database middleware and demonstrate experimentally that our approach is competitive with native implementations. Our implementation extends the GProM database middleware to support temporal queries. See https://github.com/IITDBGroup/gprom/wiki/temporal for some example usage.

In addition, our collaborators at University of Bozen have also developed a version of Postgres with native support for set-based temporal operations that is available here: http://tpg.inf.unibz.it/.

Collaborators

Anton Dignös - Free University of Bozen/Bolzano
Johann Gamper - Free University of Bozen/Bolzano
Michael H. Böhlen - University of Zurich

Publications

Snapshot Semantics for Temporal Multiset Relations
Anton Dignös, Boris Glavic, Xing Niu, Michael H. Böhlen and Johann Gamper
Proceedings of the VLDB Endowment. 12, 6 (2019) , 639–652.
```
@article{DG19,
  author = {Dign\"{o}s, Anton and Glavic, Boris and Niu, Xing and B\"{o}hlen, Michael H. and Gamper, Johann},
  journal = {Proceedings of the VLDB Endowment},
  keywords = {Temporal Databases; Annotations},
  longversionurl = {https://arxiv.org/pdf/1902.04938},
  reproducibility = {https://github.com/IITDBGroup/2019-PVLDB-Reproducibility-Snapshot-Semantics-For-Temporal-Multiset-Relations},
  projects = {Snapshot Semantics for Temporal Databases},
  number = {6},
  pages = {639--652},
  pdfurl = {http://www.vldb.org/pvldb/vol12/p639-dignoes.pdf},
  title = {{Snapshot Semantics for Temporal Multiset Relations}},
  venueshort = {PVLDB},
  volume = {12},
  year = {2019}
}
```
Snapshot semantics is widely used for evaluating queries over temporal data: temporal relations are seen as sequences of snapshot relations, and queries are evaluated at each snapshot. In this work, we demonstrate that current approaches for snapshot semantics over interval-timestamped multiset relations are subject to two bugs regarding snapshot aggregation and bag difference. We introduce a novel temporal data model based on K-relations that overcomes these bugs and prove it to correctly encode snapshot semantics. Furthermore, we present an efficient implementation of our model as a database middleware and demonstrate experimentally that our approach is competitive with native implementations.
details

Snapshot Semantics for Temporal Multiset Relations (extended version)
Anton Dignös, Boris Glavic, Xing Niu, Michael H. Böhlen and Johann Gamper
Technical Report #IIT/CS-DB-2018-03
Illinois Institute of Technology.

pdf
details

@techreport{DG18,
  author = {Dign\"{o}s, Anton and Glavic, Boris and Niu, Xing and B\"{o}hlen, Michael H. and Gamper, Johann},
  institution = {Illinois Institute of Technology},
  keywords = {Temporal Databases},
  number = {IIT/CS-DB-2018-03},
  pdfurl = {http://cs.iit.edu/%7edbgroup/assets/pdfpubls/DG18.pdf},
  title = {Snapshot Semantics for Temporal Multiset Relations (extended version)},
  venueshort = {Techreport},
  projects = {Snapshot Semantics for Temporal Databases},
  year = {2018}
}

details