AI helps you reading Science

AI generates interpretation videos

AI extracts and analyses the key points of the paper to generate videos automatically


pub
Go Generating

AI Traceability

AI parses the academic lineage of this thesis


Master Reading Tree
Generate MRT

AI Insight

AI extracts a summary of this paper


Weibo:
We propose to demonstrate MuSe, a system for Database repairs where constraints are expressed as Declarative Rules and can be interpreted in different ways by using four different semantics

MuSe: multiple deletion semantics for data repair

Hosted Content, no. 12 (2020): 2921-2924

Cited: 0|Views2244
EI
Full Text
Bibtex
Weibo

Abstract

AbstractWe propose to demonstrate MuSe, a system for Database repairs where constraints are expressed as Declarative Rules and can be interpreted in different ways by using four different semantics. Our framework may capture common, cross-relation, repair semantics such as that of SQL deletion triggers, causal rules, and denial constraint...More

Code:

Data:

0
Introduction
  • Database repair through tuple deletion has been extensively studied [2]. In particular, repairing the database using a minimum number of deletions is a desired feature [1, 4].
  • MuSe provides explanations for the repairs under the different semantics, showing the process that resulted in a specific tuple deletion.
Highlights
  • Database repair through tuple deletion has been extensively studied [2]
  • There are various approaches for database repair for different use cases and scenarios. These include repairing via integrity constraints, such as Denial Constraints (DCs) [2], and expressing dependencies between relations via causal rules [10] or SQL triggers. All of these approaches have specific semantics associated with them; namely, DCs point to a set of tuples violating a constraint, but do not specify which tuple in this set should be removed
  • Causal rules and SQL triggers point to a specific tuple that should be deleted and allow for cascade deletions. Even for the latter there is no single accepted semantics for cases where several triggers are satisfied at the same time: one approach may be to fire the triggers according to lexicographic order [9], while another is to fire them in the order in which they were created [8]
  • We can just delete the tuples ag2, ag3. This would satisfy the first constraint and the second, third and fourth constraints will be satisfied. To account for these different approaches and to provide clear and reliable semantics for database repair with deletions, we propose to demonstrate MuSe, a novel unified framework for database repair through tuple deletions that includes four different semantics of minimum repairs based on [5]
  • MuSe provides explanations for the repairs under the different semantics, showing the process that resulted in a specific tuple deletion
Results
  • In Figures 1 and 2, if the authors remove the tuples g2, a2, a3, w1, w2, p1, p2, c and add the corresponding delta tuples ∆(g2), ∆(a2), ∆(a3), ∆(w1), ∆(w2), ∆(p1), ∆(p2), ∆(c), the authors would have a stable database, as none of the rules are satisfied.
  • Step semantics: This semantics resembles SQL triggers as it allows for cascade deletion, yet it is a fine-grained semantics that evaluates one rule at a time and updates the database immediately by deleting the original tuple and adding the derived delta tuple.
  • The authors choose the assignment to rule (1) deriving ∆(a2), and update D1 so it includes the change ∆2Author = {a2}, Author2 = {a1, a3} At step 3, the authors have three satisfying assignments: to rules (1), (2), and (3).
  • Stage semantics works as follows: At the first stage, there is one assignments to rule (0) deriving ∆(g2), the authors update ∆Grants = {g2}, Grants = {g1}.
  • The authors give a brief overview of the main algorithms employed by MuSe. For stage and end semantics, the authors use strategies similar to datalog evaluation, evaluating all rules and updating the database.
  • The algorithm generates the provenance expression of each derived delta tuple and includes it in a single formula, connected by ∨ between different assignments.
  • Viewing repairs: Once the delta rules are formulated, a repair semantics has been chosen, and the participants have clicked the ‘Repair’ button, the output screen will display the deleted tuples.
  • Under the hood: To investigate the repair according to independent or step semantics, participants will click the ‘Under the Hood’ button in the output screen will show an explanation describing the deletion process, based on the repair algorithm for the chosen semantics.
Conclusion
  • Commit a repair to the database: If the participants would want to use the generated repair according to their chosen delta rules and semantics, they will click the ‘Commit’ button that will commit the repair to the database and delete the tuples shown in the output screen (Figure 4b).
  • Free exploration of MuSe: Participants will be encouraged to change the delta rules, delete some of them and add new ones, and witness the effects of their changes on the repairs under different semantics.
Funding
  • This research has been funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No 804302), the Israeli Science Foundation (ISF) Grant No 978/17, NSF awards IIS-1552538 and IIS-1703431, NIH award R01EB025021, and a Google Ph.D
Reference
  • [2] J. Chomicki and J. Marcinkowski. Minimal-change integrity maintenance using tuple deletions. Inf. Comput., 197(1-2):90–121, 2005.
    Google ScholarLocate open access versionFindings
  • [3] D. Deutch, T. Milo, S. Roy, and V. Tannen. Circuits for datalog provenance. In ICDT, pages 201–212, 2014.
    Google ScholarLocate open access versionFindings
  • [4] R. Fagin, B. Kimelfeld, and P. G. Kolaitis. Dichotomies in the complexity of preferred repairs. In PODS, pages 3–15, 2015.
    Google ScholarLocate open access versionFindings
  • [5] A. Gilad, D. Deutch, and S. Roy. On multiple semantics for declarative database repairs. In SIGMOD, pages 817–831, 2020.
    Google ScholarLocate open access versionFindings
  • [6] T. J. Green, G. Karvounarakis, and V. Tannen. Provenance semirings. In PODS, pages 31–40, 2007.
    Google ScholarLocate open access versionFindings
  • [10] S. Roy and D. Suciu. A formal approach to finding explanations for database queries. In SIGMOD, pages 1579–1590, 2014.
    Google ScholarLocate open access versionFindings
0
Your rating :

No Ratings

Tags
Comments
数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn