View publication
Title | Towards General Representability in Knowledge Exchange |
Authors | Marcelo Arenas, Jorge Pérez, Emanuel Sallinger |
Publication date | 2013 |
Abstract | In data exchange, one is typically given a source database instance and a mapping between a source schema and a target schema. The goal is then to materialize a target database instance that corresponds to the source instance and the mapping. Knowledge bases consist of explicit knowledge - as in database instances - and implicit knowledge, usually given in the form of rules. Knowledge exchange can thus be seen as a generalization of data exchange, where one is given a source knowledge base and a mapping between the source and the target schema. The goal is to materialize a target knowledge base, that is, both explicit and implicit knowledge. \n\n It was found to be desirable to consider knowledge exchange as a two-stage process, where the first step is to materialize target implicit knowledge, given only the source implicit knowledge and the source-to-target mapping. We can thus say that such target implicit knowledge "represents" the source implicit knowledge under a given mapping. Two notions of "representation" were introduced recently: safety and Q-representability. \n\n In this work, we introduce the very broad notion of general representability that captures safety and lays the foundation to the study of representability in a broad setting, in particular with the goal of extending our work to Q-representability. To achieve this, general representability is not limited to knowledge exchange, but based on arbitrary schema mappings. As an essential tool for this study, we develop conditions on schema mapping languages under which notions of equivalence important to our investigation coincide. |
Downloaded | 6 times |
Pages | 1-7 |
Conference name | Alberto Mendelzon International Workshop on Foundations of Data Management |
Publisher | CEUR Publications |
Reference URL |