Hippo: A System for Computing Consistent Answers to a Class of

Download Report

Transcript Hippo: A System for Computing Consistent Answers to a Class of

Hippo
Jan Chomicki
University at Buffalo
Jerzy Marcinkowski
Wroclaw University
Slawomir Staworko
University at Buffalo
a System for Computing
Consistent Query Answers
to a Class of SQL Queries
Motivation - Inconsistent data
Enforcing data consistency no longer applicable:
 Data Integration – Consistent data sources, but
inconsistent global view.
 Long-running transactions.
 Efficiency reasons.
Consistent Query Answers
Repair
 Instance satisfying the constraint.
 The set of changes is minimal.
There can be an exponential number of repairs.
Tuple t is a consistent answer to Q if t is an
answer to Q in every repair.
Computing CQA

Query rewriting
For query Q construct Q’ which evaluation returns
consistent answers of Q.

Logic programming
Use disjunctive program to specify repairs and
query result.

Condensed representations of repairs
Conflict Hypergraphs


Vertex – database tuple
Edge – conflicting tuples
Name
Town
J. Smith
Buffalo
J. Smith
Chicago
D. Gibs
Buffalo
M. Adams
Buffalo
M. Adams
Chicago
M. Adams
New York
Repair – Maximal Independent Set
(J.S.,BUF)
(J.S.,CHO)
(D.G.,BUF)
(M.A.,BUF)
(M.A.,CHO)
(M.A.,NYC)
Hippo – System Description


Conflict hypergraph – stored in RAM
Denial integrity constraints
[ R1 (t1 )    Rn (tn )   ]



Queries:  , /,, and pseudo  
SQL frontend – RDMBS independent
Platform independent (Java2)
Hippo is fast


Selection and Join – as fast as underlying
database system.
(QR takes approx. twice the time)
Union and Difference – takes approx. Twice
the time of simple query evaluation.
(QR the same for difference).
Future Work

Projection
–
–
–

In general problem is co-NP-data-complete.
Find an efficient heuristic.
Characterize hypergraphs where projection is
easy
Preferences
–
–
User provides preferences on resolving conflicts.
Computing Preferred CQA still easy.
References
1.
2.
M. Arenas, L. Bertossi, J. Chomicki.
Consistent Query Answers in Inconsistent
Databases. PODS’99
J. Chomicki, J. Marcinkowski. Minimal Change
Integrity Maintenance using Tuple Deletions. Under
revision for Information and Computation.
3.
J. Chomicki, J. Marcinkowski, S. Staworko.
Computing Consistent Query Answers using
Conflict Hypergraphs. Under conference submission.
4.
http://www.cse.buffalo.edu/~chomicki