Ontology matching (OM) enables semantic interoperability between different ontologies and resolves their conceptual heterogeneity by aligning related entities. OM systems currently have two prevailing design paradigms: conventional knowledge-based expert systems and newer machine learning-based predictive systems. While large language models (LLMs) and LLM agents have revolutionised data engineering and have been applied creatively in many domains, their potential for OM remains underexplored. This study introduces a novel agent-powered LLM-based design paradigm for OM systems. With consideration of several specific challenges in leveraging LLM agents for OM, we propose a generic framework, namely Agent-OM (Agent for Ontology Matching), consisting of two Siamese agents for retrieval and matching, with a set of OM tools. Our framework is implemented in a proof-of-concept system. Evaluations of three Ontology Alignment Evaluation Initiative (OAEI) tracks over state-of-the-art OM systems show that our system can achieve results very close to the long-standing best performance on simple OM tasks and can significantly improve the performance on complex and few-shot OM tasks.

Enterprise and public Knowledge Graphs (KGs) are known to be incomplete. Methods for automatic completion, sometimes by rule learning, scale well. While previous rule-based methods learn closed (non-existential) rules, we introduce Open Path (OP) rules that are constrained existential rules. We present a novel algorithm, OPRL, for learning OP rules. Closed rules complete a KG by answering queries of unclear origin, usually derived from a holdback test set in experimental settings. However, OP rules can generate relevant queries for KG completion. OPRL generates queries even when there is no closed rule to answer the query, or when the correct answer is a missing entity that is not present in the KG. For OPRL to scale well, we propose a novel embedding-based fitness function to efficiently estimate rule quality. Additionally, we introduce a novel, efficient vector computation to formally assess rule quality. We evaluate OPRL using adaptations of Freebase, YAGO2, Wikidata, and a synthetic Poker KG. We find that OPRL mines hundreds of accurate rules from massive KGs with up to 8 M facts. The OP rules generate queries with precision as high as 98% and recall of 62% on a complete KG, demonstrating the first solution for active knowledge graph completion.

Extraction of categorised named entities from text is a complex task given the availability of a variety of Named Entity Recognition (NER) models and the unstructured information encoded in different source document formats. Processing the documents to extract text, identifying suitable NER models for a task, and obtaining statistical information is important in data analysis to make informed decisions. This paper presents\footnoteThe manuscript follows guidelines to showcase a demonstration that introduces an overview of how the toolkit works: input document set, initial settings, processing, and output set. The input document set is artificial in order to show various toolkit capabilities. TNNT, a toolkit that automates the extraction of categorised named entities from unstructured information encoded in source documents, using diverse state-of-the-art (SOTA) Natural Language Processing (NLP) tools and NER models.TNNT integrates 21 different NER models as part of a Knowledge Graph Construction Pipeline (KGCP) that takes a document set as input and processes it based on the defined settings, applying the selected blocks of NER models to output the results. The toolkit generates all results with an integrated summary of the extracted entities, enabling enhanced data analysis to support the KGCP, and also, to aid further NLP tasks.

Knowledge Bases (KBs) are widely used as one of the fundamental components in Semantic Web applications as they provide facts and relationships that can be automatically understood by machines. Curated knowledge bases usually use Resource Description Framework (RDF) as the data representation model. To query the RDF-presented knowledge in curated KBs, Web interfaces are built via SPARQL Endpoints. Currently, querying SPARQL Endpoints has problems like network instability and latency, which affect the query efficiency. To address these issues, we propose a client-side caching framework, SPARQL Endpoint Caching Framework (SECF), aiming at accelerating the overall querying speed over SPARQL Endpoints. SECF identifies the potential issued queries by leveraging the querying patterns learned from clients’ historical queries and prefecthes/caches these queries. In particular, we develop a distance function based on graph edit distance to measure the similarity of SPARQL queries. We propose a feature modelling method to transform SPARQL queries to vector representation that are fed into machine-learning algorithms. A time-aware smoothing-based method, Modified Simple Exponential Smoothing (MSES), is developed for cache replacement. Extensive experiments performed on real-world queries showcase the effectiveness of our approach, which outperforms the state-of-the-art work in terms of the overall querying speed.

This paper presents a generic approach to integrate environmental sensor data efficiently, allowing the detection of relevant situations and events in near real-time through continuous querying. Data variety is addressed with the use of the Semantic Sensor Network ontology for observation data modelling, and semantic annotations for environmental phenomena. Data velocity is handled by distributing sensor data messaging and serving observations as RDF graphs on query demand. The stream processing engine presented in the paper, morph-streams++, provides adapters for different data formats and distributed processing of streams in a cluster. An evaluation of different configurations for parallelization and semantic annotation parameters proves that the described approach reduces the average latency of message processing in some cases.

This paper tackles the key scheduling problem of reducing the overall wait time of mobile clients in wireless data broadcast systems. It is observed that in periodic broadcast, new mobile clients may join in and existing mobile clients may leave anytime; in on-demand broadcast, high uplink communication cost may occur as all clients have to submit their queries every time. These are likely to degrade existing broadcasting approaches. In this work, we study the scheduling problem of XML data broadcast in a hybrid mode, where the system supports both periodic broadcast and on-demand broadcast services at the same. By taking the structural similarity between XML documents into account, only a small portion of mobile clients would be involved in the scheduling process and all mobile clients can be served more effectively. In this way, communication cost at the client side can be reduced greatly. A formal theoretical analysis of the proposed technique is presented. Based on the analysis, a novel clustering-based scheduling algorithm is developed. Moreover, we utilize an aging method to predict the distribution of incoming queries based on small samples of queries from mobile clients. Finally, we evaluate the approach through a set of experiments and the results show that it can significantly improve access efficiency for mobile clients.

Cross-validation is a commonly used method for evaluating the effectiveness of Support Vector Machines (SVMs). However, existing SVM cross-validation algorithms are not scalable to large datasets because they have to (i) hold the whole dataset in memory and/or (ii) perform a very large number of kernel value computation. In this paper, we propose a scheme to dramatically improve the scalability and efficiency of SVM cross-validation through the following key ideas. (i) To avoid holding the whole dataset in the memory and avoid performing repeated kernel value computation, we precompute the kernel values and reuse them. (ii) We store the precomputed kernel values to a high-speed storage framework, consisting of CPU memory extended by solid state drives (SSDs) and GPU memory as a cache, so that reusing (i.e., Reading) kernel values takes much lesser time than computing them on-the-fly. (iii) To further improve the efficiency of the SVM training, we apply a number of techniques for the extreme example search algorithm, design a parallel kernel value read algorithm, propose a caching strategy well-suited to the characteristics of the storage framework, and parallelize the tasks on the GPU and the CPU. For datasets of sizes that existing algorithms can handle, our scheme achieves several orders of magnitude of speedup. More importantly, our scheme enables SVM cross-validation on datasets of very large scale that existing algorithms are unable to handle.

A process trace describes the processes taken in a workflow to generate a particular result. Given many process traces, each with a large amount of very low level information, it is a challenge to make process traces meaningful to different users. It is more challenging to compare two complex process traces generated by heterogenous systems and have different levels of granularity. We present CTrace, a system that (1) lets users explore the conceptual abstraction of large process traces with different levels of granularity, and (2) provides semantic comparison among traces in which both the structural and the semantic similarity are considered. The above functions are underpinned by a novel notion of multi-granularity process trace and efficient multi-granularity similarity comparison algorithms.

The Internet of Things (IoT) has recently received considerable interest from both academia and industry that are working on technologies to develop the future Internet. It is a joint and complex discipline that requires synergetic efforts from several communities such as telecommunication industry, device manufacturers, semantic Web, and informatics and engineering. Much of the IoT initiative is supported by the capabilities of manufacturing low-cost and energy-efficient hardware for devices with communication capacities, the maturity of wireless sensor network technologies, and the interests in integrating the physical and cyber worlds. However, the heterogeneity of the “Things” makes interoperability among them a challenging problem, which prevents generic solutions from being adopted on a global scale. Furthermore, the volume, velocity and volatility of the IoT data impose significant challenges to existing information systems. Semantic technologies based on machine-interpretable representation formalism have shown promise for describing objects, sharing and integrating information, and inferring new knowledge together with other intelligent processing techniques. However, the dynamic and resource-constrained nature of the IoT requires special design considerations to be taken into account to effectively apply the semantic technologies on the real world data. In this article the authors review some of the recent developments on applying the semantic technologies to IoT.

The W3C Semantic Sensor Network Incubator group (the SSN-XG) produced an OWL 2 ontology to describe sensors and observations — the SSN ontology, available at http://purl.oclc.org/NET/ssnx/ssn. The SSN ontology can describe sensors in terms of capabilities, measurement processes, observations and deployments. This article describes the SSN ontology. It further gives an example and describes the use of the ontology in recent research projects.

Due to environmental and contextual changes, it is often required to configure and re-configure sensor network services. The burden of programming a sensor network, as a consequence of these changes, can be alleviated with a plug-and-play approach that allows dynamic adaptation of sensor network services. In this paper, we briefly present a semantics-based methodology to avail self-configuration of sensor services with minimal human intervention. We also provide a practical usage scenario and our preliminary work in this context.

In order for the Bureau of Meteorology (BOM), Australia, to build and maintain an integrated national water information system, over 240 organisations are required to provide their data to BOM. These organisations use a wide range of systems and data formats. To ensure robust and reliable data delivery, BOM has established Water Data Transfer Format (WDTF) as a standard format for data transfer. Meanwhile, the Water Regulations 2008 were enacted to specify the water information required from organisations. This paper analyses semantic gaps between data from organisations, WDTF, and the Regulations requirements, and proposes a knowledge-driven approach in which these gaps are captured in a way that facilitates data translation and validation. Throughout the paper, real data examples are used to illustrate the details of the approach and its feasibility.

We investigate the problem of creating and analyzing samples of relational databases to find relationships between string-valued attributes. Our focus is on identifying attribute pairs whose value sets overlap, a pre-condition for typical joins over such attributes. However, real-world data sets are often 'dirty', especially when integrating data from different sources. To deal with this issue, we propose new similarity measures between sets of strings, which not only consider set based similarity, but also similarity between strings instances. To make the measures effective, we develop efficient algorithms for distributed sample creation and similarity computation. Test results show that for dirty data our measures are more accurate for measuring value overlap than existing sample-based methods, but we also observe that there is a clear tradeoff between accuracy and speed. This motivates a two-stage filtering approach, with both measures operating on the same samples.

Abstract: To enhance aerospace applications such as supply chain management or maintenance tracking and reliability assessment, aircraft manufacturers need to enrich their electronic documentation systems with better conceptualizations of the aerospace domain. Because of the number and diversity of products and parts in an aircraft, ontologies to model this domain can potentially be very large. Therefore, it is critical to have scalable ontology generation approaches, along with an understanding of how the design of these ontologies has an impact on the performance of description logic reasoners that can operate over them. In this paper, we investigate how to achieve this goal though the application of best practice ontology engineering patterns. In particular, we examine two types of ontology engineering patterns used to instantiate large‐scale ontologies based on part–whole relations. The first approach is a direct implementation of the part–whole guidelines published by W3C. The second approach uses right‐identity axioms supported by the EL+ description logic. The results of our empirical evaluation show the benefits of the latter approach, whereby the CEL reasoner is able to perform the task of classification over large ontologies significantly faster than the description logic reasoners FaCT++, RACER and Pellet that operate over comparable ontologies designed using the W3C pattern.

Scientists have been organising the forms of natural life into structured hierarchical systems since Linnaeus in the 18th century. Much more recently, computer scientists have developed a class of languages, called description logics (DL), that are aimed at describing concepts so that they may be automatically classified in hierarchical structures. These languages are being adopted in recent proposals for ontology definition that underly the semantic Web, particularly OWL-DL (Bechofer et al., 2003). In this paper we study the applicability of modern description logics to the application of animal systematics. We would like to improve both the process of scientific classification itself, and the methods for communication and integration of taxonomic knowledge. As a case study, we consider a published scientific treatment of Epopostruma, a genus of Australian Formicidae (ants) (Shattuck, 2000). We focus on expressing the morphological characters of Epopostruma, that is the features that derive from the form, structures, homologies and metamorphoses which characterise an individual. We express these characters in the description logic ALCQHIO/sub R//sup +/(D)/sup -/ underlying OWL-DL. Racer (Haarslev and Moller, 2001) is a readily-available reasoner ALCQHIO/sub R//sup +/(D)/sup -/, and is used in this paper to support the development of the DL application to animal systematics. We have used the native syntax of Racer for DL expressions in this paper. We find that most of the language used in a scientific description is readily adapted to the formal description logic language, with the exception of spatio-temporal elements and some higher-order constructs. We show that the reasoning capability is sufficient for consistency checking and retrieval of taxonomic knowledge. We discuss some benefits of the representation to assist the work of biological systematists.

We propose to use an explicit taxonomy as a key element in a system that integrates distributed, heterogeneous biological databases. The system features a Web interface that enables a user to navigate a taxonomy tree and to post queries to distributed biological collections databases based on a selected node in the taxonomy tree. We use the taxonomy dataset published by the National Centre for Biotechnology Information as a canonical taxonomy and we describe the representation of database coverage in terms of this taxonomy. We present some rules to detect the semantic conflicts between the canonical taxonomy and those partially represented in the independent databases. We provide facilities to resolve the conflicts. We propose a mapping algorithm and describe how mapping information is used to generate queries for distributed searching. The Web-based taxonomy browser called TaxaServer is implemented as a component of our Bioplex system.

Predictive simulation systems, embedding models of real-world processes, are usually designed as stand-alone computer packages. Users wishing to work on broader problems of extended scope need access to several models. This access can be provided by integrating several packages into a decision support system offering a coherent map-based problem-oriented interface to the under-lying models. This paper proposes the FMISDS integration approach that offers a problem-oriented method for designing the integrated system. It distinguishes syntactic and semantic issues in model integration, enabling a declarative description of the models and their conceptual relationships. The integration effort is supported by a reusable toolkit that interprets the declarative descriptions. FMISDS is demonstrated in a spatial decision support system for water catchment management.

Spatial information systems are employed to record the schematics of large networks for utilities and telecommunication organisations. Concurrent users access the spatial information systems to insert new designs, update designs and to record the current status of the network. Concurrency must be managed such that the data is not corrupted. This paper introduces a model for long duration transactions based on the use of integrity constraints. The model, called COLT (COnstraint-based Long Transaction), is domain-independent but is especially suitable for spatial information systems and domains where its intent naturally manifests constraints that relate data items. User-defined constraints and database integrity constraints are used both to specify the correctness criteria and to manage long duration ad-hoc transactions. The model is a generalisation of the traditional ACID transaction model. The approach enables the specification of the correctness criteria to be declarative and the user needs no knowledge of the semantics of other concurrently executing transactions. The COLT model is formally specified using the external actions of the I/O automata method and correctness is defined with respect to that method. Correctness is defined independent of implementation.