Reasoning on regular path queries

Abstract

Current information systems are required to deal with more complex data with respect to traditional relational data. The database community has already proposed abstractions for these kinds of data, in particular in terms of semistructured data models. A semistructured model conceives a database essentially as a finite directed labeled graph whose nodes represent objects, and whose edges represent relationships between objects. In the same way as conjunctive queries form the core of any query language for the relational model, regular path queries (RPQs) and their variants are considered the basic querying mechanisms for semistructured data.Besides the basic task of query answering, i.e., evaluating a query over a database, databases should support other reasoning services related to querying. One of the most important is query containment, i.e., verifying whether for all databases the answer to a query is a subset of the answer to a second query. Another important reasoning service that has received considerable attention in the recent years is view-based query processing, which amounts to processing queries based on a set of materialized views, rather than on the raw data in the database.The goal of this paper is to describe basic results and techniques concerning query containment and view based query processing for the class of two-way regular-path queries (which extend RPQs with the inverse operator). We will demonstrate that the basic services for reasoning about two way regular path queries are decidable, thus showing that the limited form of recursion expressible by these queries does not endanger the decidability of reasoning. Besides the specific results, our methods show the power of two-way automata in reasoning on complex queries.