Consistent Query Answers in Inconsistent Databases

Consistent Query Answers in Inconsistent Databases

1999 | Marcelo Arenas, Leopoldo Bertossi, Jan Chomicki
This paper presents a method for computing consistent query answers in relational databases that may violate integrity constraints. The approach is based on the notion of repairs, which are database instances that satisfy the integrity constraints and are minimal in terms of distance from the original database. A consistent query answer is one that is the same across all possible repairs of the database. The method involves an iterative procedure that generates a new query, denoted $ T_{\omega}(Q) $, which, when evaluated on any database, returns the set of consistent answers to the original query $ Q $. The paper introduces the concept of residues, which are derived from integrity constraints and used to iteratively modify the query. The process of generating residues involves transforming integrity constraints into a standard format and then deriving rules that help in computing the consistent answers. The iterative procedure is applied to the query to generate $ T_{\omega}(Q) $, which ensures that the answers are consistent across all possible repairs of the database. The method is proven to be sound and complete for certain classes of constraints and queries. It is also shown that the procedure terminates for several classes of constraints. The paper discusses the application of this method in various areas such as data warehousing, database integration, and active databases. In data warehousing, the method helps in identifying clean data and safely removing unclean data. In database integration, it helps in determining which query answers are consistent with the integrity constraints. In active databases, it ensures that query answers reflect the consistency of the database even during temporary inconsistencies. The paper also addresses the issue of termination, showing that the operator $ T_{\omega} $ can return a finite set of formulas under certain conditions. The results are extended to various types of constraints and queries, including functional dependencies and set inclusion dependencies. The paper concludes with a discussion of related work and future research directions in the area of consistent query answers in inconsistent databases.This paper presents a method for computing consistent query answers in relational databases that may violate integrity constraints. The approach is based on the notion of repairs, which are database instances that satisfy the integrity constraints and are minimal in terms of distance from the original database. A consistent query answer is one that is the same across all possible repairs of the database. The method involves an iterative procedure that generates a new query, denoted $ T_{\omega}(Q) $, which, when evaluated on any database, returns the set of consistent answers to the original query $ Q $. The paper introduces the concept of residues, which are derived from integrity constraints and used to iteratively modify the query. The process of generating residues involves transforming integrity constraints into a standard format and then deriving rules that help in computing the consistent answers. The iterative procedure is applied to the query to generate $ T_{\omega}(Q) $, which ensures that the answers are consistent across all possible repairs of the database. The method is proven to be sound and complete for certain classes of constraints and queries. It is also shown that the procedure terminates for several classes of constraints. The paper discusses the application of this method in various areas such as data warehousing, database integration, and active databases. In data warehousing, the method helps in identifying clean data and safely removing unclean data. In database integration, it helps in determining which query answers are consistent with the integrity constraints. In active databases, it ensures that query answers reflect the consistency of the database even during temporary inconsistencies. The paper also addresses the issue of termination, showing that the operator $ T_{\omega} $ can return a finite set of formulas under certain conditions. The results are extended to various types of constraints and queries, including functional dependencies and set inclusion dependencies. The paper concludes with a discussion of related work and future research directions in the area of consistent query answers in inconsistent databases.
Reach us at info@study.space