Ontology-mediated query answering (OMQA) emerged as a paradigm to enhance querying of data sources with an ontology that encodes background knowledge. In applications involving large amounts of data from multiple data sources, it might well be the case that inconsistency arises, making standard query answering useless, since everything is entailed by an inconsistent knowledge base. Being able to provide meaningful query answers in the presence of inconsistency is thus a critical issue to make OMQA systems successful in practice. The problem of querying inconsistent knowledge has attracted a great deal of interest over the years. Different inconsistency-tolerant semantics of query answering have been proposed, that is, approaches to answer queries in a meaningful way despite the knowledge at hand being inconsistent. Most of the semantics in the literature are based on the notion of repair, that is, a “maximal” consistent subset of the database. In general, there can be several repairs, so it is often natural and desirable to express preferences among them. In this paper, we propose a framework for querying inconsistent knowledge bases under user preferences for existential rule languages. Specifically, we introduce preference rules, a declarative formalism which enable users to express (i) preferences over both the database and the knowledge that can be derived from it via an ontology, and (ii) preconditions for preferences to hold. We then define two notions of preferred repairs which take preference rules into account. This naturally leads us to introducing preference-aware counterparts of popular inconsistency-tolerant semantics, where only preferred repairs are considered for query answering. We provide a thorough analysis of the data and combined complexity of different relevant problems for a wide range of existential rule languages.

本体介导的查询回答（OMQA）作为一种范式出现，可以通过编码背景知识的本体来增强对数据源的查询。在涉及来自多个数据源的大量数据的应用程序中，很可能会出现不一致的情况，从而使标准查询回答变得无用，因为一切都由不一致的知识库所包含。因此，能够在存在不一致的情况下提供有意义的查询答案是使 OMQA 系统在实践中成功的关键问题。多年来，查询不一致知识的问题引起了极大的兴趣。已经提出了不同的查询回答的不一致性容忍语义，即尽管手头的知识不一致，但仍以有意义的方式回答查询的方法。repair，即数据库的“最大”一致子集。通常，可能有多个修复，因此在它们之间表达偏好通常是自然且可取的。在本文中，我们提出了一个框架，用于在存在规则语言的用户偏好下查询不一致的知识库。具体来说，我们引入了偏好规则，这是一种声明性形式，它使用户能够表达（i）对数据库和可以通过本体派生的知识的偏好，以及（ii）保持偏好的先决条件。然后，我们定义了两个首选维修的概念其中考虑了偏好规则。这自然导致我们引入流行的不一致性容忍语义的偏好感知对应物，其中只考虑首选修复来进行查询回答。我们为广泛的存在规则语言提供了对数据和不同相关问题的综合复杂性的全面分析。