Vast amounts of text on the Web are unstructured and ungrammatical, such as classified ads, auction listings, forum postings, etc. We call such text posts. Despite their inconsistent structure and lack of grammar, posts are full of useful information. This paper presents work on semi-automatically building tables of relational information, called reference sets, by analyzing such posts directly. Reference sets can be applied to a number of tasks such as ontology maintenance and information extraction. Our reference-set construction method starts with just a small amount of background knowledge, and constructs tuples representing the entities in the posts to form a reference set. We also describe an extension to this approach for the special case where even this small amount of background knowledge is impossible to discover and use. To evaluate the utility of the machine-constructed reference sets, we compare them to manually constructed reference sets in the context of reference-set-based information extraction. Our results show the reference sets constructed by our method outperform manually constructed reference sets. We also compare the reference-set-based extraction approach using the machine-constructed reference set to supervised extraction approaches using generic features. These results demonstrate that using machine-constructed reference sets outperforms the supervised methods, even though the supervised methods require training data.

This article introduces the challenge of digital preservation in the area of engineering design and manufacturing and presents a methodology to apply knowledge representation and semantic techniques to develop Digital Engineering Archives. This work is part of an ongoing, multiuniversity, effort to create cyber infrastructure-based engineering repositories for undergraduates (CIBER-U) to support engineering design education. The technical approach is to use knowledge representation techniques to create formal models of engineering data elements, workflows and processes. With these formal engineering knowledge and processes can be captured and preserved with some guarantee of long-term interpretability. The article presents examples of how the techniques can be used to encode specific engineering information packages and workflows. These techniques are being integrated into a semantic wiki that supports the CIBER-U engineering education activities across nine universities and involving over 3500 students since 2006.

The aim of semantic science is to allow for the publications of ontologies, observation data, and hypotheses/theories. Hypotheses make predictions on data and on new cases. Those hypotheses that fit the available evidence are called theories. This paper considers how thoeries can be used for predictions in new cases. Theories are typically very narrow and not all of the inputs to a theory are observed, so to make predictions on a particular case, many theories need to be used. Without any global design, the available theories do not necessarily fit together nicely. This paper explains how theories can be combined into theory ensembles to make predictions on a particular case. This is needed to evaluate theories, and to make useful predictions. We motivate and give desiderata for theory ensembles for level 1, feature-based, semantic science, which assumes that the data and the theories can be described in terms of features (random variables).

Efficiently extracting a module from a given ontology that captures all the ontology's knowledge about a set of specified terms is well-understood task. It can be solved, for instance, by locality-based modules. In contrast, extracting all modules of an ontology is computationally difficult because there can be exponentially many. However, it is reasonable to assume that, by revealing the modular structure of an ontology, we can obtain information about its topicality, connectedness, structure, superfluous parts, or agreement between actual and intended modeling. Furthermore, incremental reasoning makes use of a number of, although not all possible, modules of an ontology. Chances are that real-life ontologies have significantly fewer modules than the worst cases. We report on experiments to obtain or estimate this number and to evaluate the modular structure of an ontology where we succeeded to compute it. In that evaluation, we look at the number and sizes of the modules, as well as the relation between module sizes and number and sizes of signatures that lead to the module.

Even in small ontologies that only contain Figure 1: A justification for Person tens of axioms, there can be multiple reasons for an entailment, none of which may be obvious. It is for this a number of justifications that all participants ranked "difficult" reason that there has recently been a lot of focus on generating to "impossible" to understand. This includes people explanations for entailments in ontologies. In the who have over two years experience of working with OWL, OWL world, justifications are a popular form of explanation building ontologies and even includes people who have developed for entailments. A justification is a minimal subset OWL reasoners. This is indicative that justification of an ontology that is sufficient for an entailment to hold understanding can be a real problem.

Reasoning with dates and duration has long been addressed by the community. Existing duration ontologies, however, lack complete axiomatizations of their intended models; many simply represent timedurations as real numbers and treat the duration function as a metric on the timeline. We show that such approaches are inadequate and provide a first-order ontology of duration that overcomes these limitations.

We propose a new family of probabilistic description logics (DLs) that, in contrast to most existing approaches, are derived in a principled way from Halpern's probabilistic first-order logic. The resulting probabilistic DLs have a two-dimensional semantics similar to certain popular combinations of DLs with temporal logic and are well-suited for capturing subjective probabilities. Our main contribution is a detailed study of the complexity of reasoning in the new family of probabilistic DLs, showing that it ranges from PTime for weak variants based on the lightweight DL EL to undecidable for some expressive variants based on the DL ALC.

The DL-Lite family of Description Logics has been designed with the specific goal of allowing for answering complex queries (in particular, conjunctive queries) over ontologies with very large instance sets (ABoxes). So far, in DL-Lite systems, this goal has been actually achieved only for relatively simple (short) conjunctive queries. In this paper we present Presto, a new query answering technique for DL-Lite ontologies, and an experimental comparison of Presto with the main previous approaches to query answering in DL-Lite. In practice, our experiments show that, in real ontologies, current techniques are only able to answer conjunctive queries of less than 7-10 atoms (depending on the complexity of the TBox), while Presto is actually able to handle conjunctive queries of up to 30 atoms. Furthermore, in the cases that are already successfully handled by previous approaches, Presto is significantly more efficient.

Databases and related information systems can benefit from the use of ontologies to enrich the data with general background knowledge. The DL-Lite family of ontology languages was specifically tailored towards such ontology-based data access, enabling an implementation in a relational database management system (RDBMS) based on a query rewriting approach. In this paper, we propose an alternative approach to implementing ontology-based data access in DL-Lite. The distinguishing feature of our approach is to allow rewriting of both the query and the data. We show that, in contrast to the existing approaches, no exponential blowup is produced by the rewritings. Based on experiments with a number of real-world ontologies, we demonstrate that query execution in the proposed approach is often more efficient than in existing approaches, especially for large ontologies. We also show how to seamlessly integrate the data rewriting step of our approach into an RDBMS using views (which solves the update problem) and make an interesting observation regarding the succinctness of queries in the original query rewriting approach.

Recent years have seen the advent of large and complex ontologies, most notably in the medical domain. As a consequence, structuring mechanisms for ontologies are nowadays viewed as an indispensible tool. A basic such mechanism is the automatic decomposition of the vocabulary of an ontology into independent parts. In this paper, we study decompositions that are syntax independent in the sense that the resulting partitioning depends only on the meaning of the vocabulary items, but not on the concrete syntactic form of the axioms in the ontology. We present the first systematic investigation of decompositions of this type in the context of ontologies. Specifically, we focus on ontologies formulated in description logics and provide a variety of results that range from theorems stating the existence of unique finest decompositions to complexity results and algorithms computing decompositions. We also investigate the relationship between the existence of unique finite decompositions and a variant of the Craig interpolation property called parallel interpolation.