Publications

2022

1. Declarative RDF graph generation from heterogeneous (semi-)structured data: A systematic literature review Van Assche Dylan, Delva Thomas, Haesendonck Gerald, Heyvaert Pieter, De Meester Ben, and Dimou Anastasia Journal of Web Semantics 2022 [Abstract] [DOI] [PDF]

   More and more data in various formats are integrated into knowledge graphs. However, there is no overview of existing approaches for generating knowledge graphs from heterogeneous (semi-)structured data, making it difficult to select the right one for a certain use case. To support better decision making, we study the existing approaches for generating knowledge graphs from heterogeneous (semi-)structured data relying on mapping languages. In this paper, we investigated existing mapping languages for schema and data transformations, and corresponding materialization and virtualization systems that generate knowledge graphs. We gather and unify 52 articles regarding knowledge graph generation from heterogeneous (semi-)structured data. We assess 15 characteristics on mapping languages for schema transformations, 5 characteristics for data transformations, and 14 characteristics for systems. Our survey paper provides an overview of the mapping languages and systems proposed the past two decades. Our work paves the way towards a better adoption of knowledge graph generation, as the right mapping language and system can be selected for each use case.

2. Continuous generation of versioned collections’ members with RML and LDES Van Assche Dylan, Min Oo Sitt, Meléndez Julián Rojas, and Colpaert Pieter In Proceedings of the 3rd International Workshop on Knowledge Graph Construction (KGCW 2022) co-located with 19th Extended Semantic Web Conference (ESWC 2022) 2022 [Abstract] [URN] [PDF] [Slides]

   When evolving datasets are used to generate a knowledge graph, it is usually challenging to keep the graph synchronized in a timely manner when changes occur in the source data. Current approaches fully regenerate a knowledge graph in such cases, which may be time consuming depending on the data type, size, and update frequency. We propose a continuous knowledge graph generation approach that can be applied on different types of data sources. We describe continuously updating knowledge graph versions represented as a Linked Data Events Stream, and use an RML processor for RDF generation. In this paper, we present our approach and demonstrate it on different types of data such as bike-sharing, public transport timetables, and weather data. By describing entities with unique, immutable, and reproducible IRIs, we were able to identify changes in the original data collection, reducing the number of materialized triples and generation time. Our use-cases show the importance of mechanisms to derive unique and stable IRI strategies of data source updates, to enable efficient knowledge graph generation pipelines. In the future, we will extend our approach to handle deletions in data collections, and conduct an extensive performance evaluation.

3. Balancing RDF generation from heterogeneous data sources Van Assche Dylan In The Semantic Web: ESWC 2022 Satellite Events 2022 [Abstract] [PDF] [Poster] [Slides]

   Knowledge graphs in RDF are often generated from heterogeneous data sources to power services. However, knowledge graph generation is an unbalanced effort for producers compared to consumers of a knowledge graph. In this paper, I present my research about (i) investigating current RDF knowledge graph production and consumption approaches, and (ii) how to involve the consumer into a hybrid RDF generation approach to reduce the necessary resources for generating RDF for producers and consumers. I discuss the shortcomings of existing approaches for RDF generation from heterogeneous data sources (i.e., materialization and virtualization) and how I will address these: a Systematic Literature Review; an analysis and a set of guidelines for producers to select the right approach for an use case; and a combined hybrid approach to balance the producer’s and consumer’s effort in RDF generation. I already performed a Systematic Literature Review to get an overview of the existing approaches for RDF production from heterogeneous data sources. These results will be used to establish a set of producer guidelines, a benchmark to compare the current materialization and virtualization approaches, and evaluate the proposed hybrid approach. Thanks to my research, knowledge graph production and consumption will be more balanced and accessible to smaller companies and individuals. This way, they can focus on providing better services on top of a knowledge graph instead of being limited by the lack of computing resources to harvest enormous amounts of data from the Web and integrate it into a knowledge graph.

2021

1. RML2SHACL: RDF generation taking shape Delva Thomas, Min Oo Sitt, Van Assche Dylan, Lieber Sven, and Dimou Anastasia In K-CAP ’21: Proocedings of the 11th on Knowledge Capture Conference 2021 [Abstract] [PDF]

   RDF graphs are often generated by mapping data in other (semi-)structured data formats to RDF. Such mapped graphs have a repetitive structure defined by (i) the mapping rules and (ii) the schema of the input sources. However, this information is not exploited beyond its original scope. SHACL was recently introduced to model constraints that RDF graphs should validate. The shapes and their constraints are either manually defined or derived from ontologies or RDF graphs. We investigate a method to derive the shapes and their constraints from mapping rules, allowing the generation of the RDF graph and the corresponding shapes in one step. In this paper, we present RML2SHACL: an approach to generate SHACL shapes that validate RDF graphs defined by RML mapping rules. RML2SHACL relies on our proposed set of correspondences between RML and SHACL constructs. RML2SHACL covers a large variety of RML constructs, as proven by generating shapes for the RML test cases. A comparative analysis shows that shapes generated by RML2SHACL are similar to shapes generated by ontology-based tools, with a larger focus on data value-based constraints instead of schema-based constraints. We also found that RML2SHACL has a faster execution time than data-graph based approaches for data sizes of 90MB and higher.

2. Towards a more human-friendly knowledge graph generation & publication Van Assche Dylan, Delva Thomas, Heyvaert Pieter, De Meester Ben, and Dimou Anastasia In International Semantic Web Conference (ISWC) 2021: Posters, Demos, and Industry Tracks 2021 [Abstract] [URN] [PDF] [Video] [Demo]

   Human-friendly representations of mapping languages, e.g.,YARRRML or ShExML, allow seamlessly integrating heterogeneous data into knowledge graphs. However, they do not describe how a generated knowledge graph should be exported or published, leaving the knowledge graph generation and publication disconnected. To address this, we recently introduced the Logical Target in RML, but an alignment with the human-friendly representations of mapping languages is still pending. In this demo paper, we (i) align RML’s Logical Target with YARRRML to provide a more human-friendly syntax for knowledge graph’s generation and publication and (ii) demonstrate our approach in YARRRML’s editor Matey. YARRRML users are now able to describe how their generated knowledge graphs are exported to one or multiple targets, paving the path towards a more reproducible and human-friendly workflow for generating and publishing knowledge graphs.

3. Leveraging semantic technologies for digital interoperability in the European Railway domain Best Paper Award Meléndez Julián Rojas, Aguado Marina, Vasilopoulou Polynmia, Velitchkov Ivo, Van Assche Dylan, Colpaert Pieter, and Verborgh Ruben In The Semantic Web - ISWC 2021 2021 [Abstract] [HTML] [DOI]

   The European Union Agency for Railways acts as an European authority, tasked with the provision of a legal and technical framework to support harmonized and safe cross-border railway operations throughout the EU. So far, the agency relied on traditional application-centric approaches to support the data exchange among multiple actors interacting within the railway domain. This lead however, to isolated digital environments that consequently added barriers to digital interoperability while increasing the cost of maintenance and innovation. In this work, we show how Semantic Web technologies are leveraged to create a semantic layer for data integration across the base registries maintained by the agency. We validate the usefulness of this approach by supporting route compatibility checks, a highly demanded use case in this domain, which was not available over the agency’s registries before. Our contributions include (i) an official ontology for the railway infrastructure and authorized vehicle types, including 28 reference datasets; (ii) a reusable Knowledge Graph describing the European railway infrastructure; (iii) a cost-efficient system architecture that enables high-flexibility for use case development; and (iv) an open source and RDF native Web application to support route compatibility checks. This work demonstrates how data-centric system design, powered by Semantic Web technologies and Linked Data principles, provides a framework to achieve data interoperability and unlock new and innovative use cases and applications. Based on the results obtained during this work, ERA officially decided to make Semantic Web and Linked Data-based approaches, the default setting for any future development of the data, registers and specifications under the agency’s remit for data exchange mandated by the EU legal framework. The next steps, which are already underway, include further developing and bringing these solutions to a production-ready state.

4. BESOCIAL: A Sustainable Knowledge Graph-based Workflow for Social Media Archiving Lieber Sven, Van Assche Dylan, Chambers Sally, Messens Fien, Geeraert Friedel, Birkholz M. Jullie, and Dimou Anastasia In Further with Knowledge Graphs 2021 [Abstract] [HTML] [PDF]

   Social media as infrastructure for public discourse provide valuable information that needs to be preserved. Several tools for social media harvesting exist, but still only fragmented workflows may be formed with different combinations of such tools. On top of that, social media data but also preservation-related metadata standards are heterogeneous, resulting in a costly manual process. In the framework of BESOCIAL at the Royal Library of Belgium (KBR), we develop a sustainable social media archiving workflow that integrates heterogeneous data sources in a Europeana and PREMIS-based data model to describe data preserved by open source tools. This allows data stewardship on a uniform representation and we generate metadata records automatically via queries. In this paper, we present a comparison of social media harvesting tools and our Knowledge Graph-based solution which reuses off-the-shelf open source tools to harvest social media and automatically generate preservation-related metadata records. We validate our solution by generating Encoded Archival Description (EAD) and bibliographic MARC records for preservation of harvested social media collections from Twitter collected at KBR. Other archiving institutions can build upon our solution and customize it to their own social media archiving policies.

5. Target in RML Van Assche Dylan, and Dimou Anastasia Technical Report 2021 [Abstract] [HTML]

   This document describes a formal representation to align the RDF Mapping Language (RML) with Target (DataIO) to describe in RML mapping rules how a generated Knowledge Graph should be directed and exported. The current document describes how Target is integrated with RML through definitions and examples.

6. DataIO Van Assche Dylan, and Dimou Anastasia Technical Report 2021 [Abstract] [HTML]

   This document describes Target, a formal model and a common representation for specifying how a Knowledge Graph should be exported to a given target. Target reuses existing data access descriptions and is therefore not limited to a specific set of targets. The current document describes the Target concepts through definitions and examples.

7. Integrating nested data into knowledge graphs with RML fields Delva Thomas, Van Assche Dylan, Heyvaert Pieter, De Meester Ben, and Dimou Anastasia In Proceedings of the 2nd International Workshop on Knowledge Graph Construction 2021 [Abstract] [HTML] [PDF]

   To support business decisions or improve operational efficiency, heterogeneous data is often integrated into a knowledge graph. This integration can be achieved with one of the existing declarative mapping languages, which offer declarative data integration in the form of knowledge graphs. However, current mapping languages cannot always integrate data with nested structure, such as JSON or XML files or JSON documents stored in a database column. We designed a backwards-compatible extension of the RDF Mapping Language (RML) which empowers it to integrate nested data: RML fields. In this paper, we introduce RML fields, compare it with the state of the art in mapping languages, and validate it on mapping challenges formulated by the Knowledge Graph Construction W3C community group. Our extension allows to address several of the challenges related to nested data that were previously not possible. RML fields can be used to integrate even more datasets into knowledge graphs with all the advantages of using a language specially designed for that purpose. Our extension currently is intended to integrate multiple data sets independently, but some use cases require joins or other operations during knowledge graph generation, which we will investigate in the future.

8. Leveraging Web of Things W3C recommendations for knowledge graphs generation Van Assche Dylan, Haesendonck Gerald, Gertjan De Mulder, Delva Thomas, Heyvaert Pieter, De Meester Ben, and Dimou Anastasia In Proceedings of the 21st International Conference on Web Engineering 2021 [Abstract] [HTML] [DOI] [PDF] [Slides] [Video]

   Constructing a knowledge graph with mapping languages, such as RML or SPARQL-Generate, allows seamlessly integrating heterogeneous data by defining access-specific definitions for e.g., databases or files. However, such mapping languages have limited support for describing Web APIs and no support for describing data with varying velocities, as needed for e.g., streams, neither for the input data nor for the output RDF. This hampers the smooth and reproducible generation of knowledge graphs from heterogeneous data and their continuous integration for consumption since each implementation provides its own extensions. Recently, the Web of Things (WoT) Working Group released a set of recommendations to provide a machine-readable description of metadata and network-facing interfaces for Web APIs and streams. In this paper, we investigated (i) how mapping languages can be aligned with the newly specified recommendations to describe and handle heterogeneous data with varying velocities and Web APIs, and (ii) how such descriptions can be used to indicate how the generated knowledge graph should be exported. We extended RML’s Logical Source to support WoT descriptions of Web APIs and streams, and introduced RML’s Logical Target to describe the generated knowledge graph reusing the same descriptions. We implemented these extensions in the RMLMapper and RMLStreamer, and validated our approach in two use cases. Mapping languages are now able to use the same descriptions to define the input data but also the output RDF. This way, our work paves the way towards more reproducible workflows for knowledge graph generation.

2020

1. Efficient Live Public Transport Data Sharing for Route Planning on the Web Meléndez Julián Rojas, Van Assche Dylan, Delva Harm, Colpaert Pieter, and Verborgh Ruben In Proceedings of the 20th International Conference on Web Engineering 2020 [Abstract] [HTML] [DOI]

   Web-based information services transformed how we interact with public transport. Discovering alternatives to reach destinations and obtaining live updates about them is necessary to optimize journeys and improve the quality of travellers’ experience. However, keeping travellers updated with opportune information is demanding. Traditional Web APIs for live public transport data follow a polling approach and allocate all data processing on either data providers, lowering data accessibility, or data consumers, increasing the costs of innovative solutions. Moreover, data processing load increases further because previously obtained route plans are fully recalculated when live updates occur. In between solutions sharing processing load between clients and servers, and alternative Web API architectures were not thoroughly investigated yet. We study performance trade-offs of polling and push-based Web architectures to efficiently publish and consume live public transport data. We implement (i) alternative architectures that allow sharing data processing load between clients and servers, and evaluate their performance following polling- and push-based approaches; (ii) a rollback mechanism that extends the Connection Scan Algorithm to avoid unnecessary full route plan recalculations upon live updates. Evaluations show polling as a more efficient alternative on CPU and RAM but hint towards push-based alternatives when bandwidth is a concern. Clients update route plan results 8–10 times faster with our rollback approach. Smarter API design combining polling and push-based Web interfaces for live public transport data reduces the intrinsic costs of data sharing by equitably distributing the processing load between clients and servers. Future work can investigate more complex multimodal transport scenarios.

2019

1. Publiceren van updates over openbaarvervoersdata Van Assche Dylan Master Thesis 2019 [Abstract] [HTML] [PDF]

   Publishing real time public transport data can be a heavy task for Open Data publishers (RPC APIs) and consumers (data dumps). With Linked Connections we try to find a compromise between both worlds. Since Linked Connections is not optimised for real time data yet, we investigate how the Linked Connections server can publish real time data in a cost-efficient way. At the same time, we examine how we can optimise the algorithms to use Linked Connections on the clients. We propose a list of changes to these algorithms to reduce the processing time and computing resources. By using a separate real time resource on the Linked Connections server, only the updates to the data must be transferred. We want to reduce further the necessary resources for the client and the server for Linked Connections. We achieve this goal by modifying the Connection Scan Algorithm (CSA) and liveboard algorithms on the client. We reduced the processing time of a journey with a factor of 2 - 3 using the CSA. A liveboard could be updated 10 times faster than in the original implementation. The amount of network bandwidth is reduced by 15 % for the CSA and by 21 % for the liveboard algorithm. Based on these results, we conclude that a publish-subscribe approach (Server-Sent-Events) is more efficient than a polling approach (HTTP polling) for real time Open Data. Using a real time resource, the efficiency of the client’s algorithms is significantly improved.