Workshops & Tutorials (June 6-7)

Organisers: Cogan Shimizu and Karl Hammar

Modular ontology modeling (MOMo) using ontology design patterns enables non-experts to develop ontologies with reasonable degree of correctness and efficiency. However, the tooling required to fully realize pattern-based approaches to ontology development, have long been lacking; consequently, comparatively few ontology engineering projects have in fact been carried out using this promising approach. The authors have developed what we believe is the first graphical drag-and-drop-based tool for modular pattern-based ontology engineering, CoModIDE. During this tutorial, we will demonstrate CoModIDE and teach the participants to use it. This tutorial is aimed at both novice ontologists (who may directly benefit from having this easy-to-use and freely available tool available, in their own modelling work) and more experienced ontology engineering researchers (who may find the tool useful as an aid in method development and evaluation, or as a teaching aid when introducing ontology engineering to less experienced colleagues).

Website: https://comodide.com/tutorial.html

Video: https://comodide.com/momo-comodide-intro-video.mp4

Organisers: Sarra Ben Abbès, Rim Hantach and Philippe Calvez

In recent years, deep learning is applied successfully and achieved state-of-the-art performance in a variety of domains, such as image analysis. Despite this success, deep learning models remain hard to analyse data and understand what knowledge is represented in them, and how they generate decisions.
Deep Learning (DL) meets Natural Language Processing (NLP) to solve human language problems for further applications such as information extraction, machine translation, search and summarization. Previous works have attested the positive impact of domain knowledge on data analysis and vice versa, for example pre-processing data, searching data, redundancy and inconsistency data, knowledge engineering, domain concepts and relationships extraction, etc. Ontology is a structured knowledge representation that facilitates data access (data sharing and reuse) and assists the DL process as well. DL meets recently ontologies and tries to model data representations with many layers of non-linear transformations.
The combination of DL, ontologies and NLP might be beneficial for different tasks:
1. Deep Learning for Ontologies: ontology population, ontology extension, ontology learning, ontology alignment and integration,
2. Ontologies for Deep Learning: semantic graph embeddings, latent semantic rep-resentation, hybrid embeddings (symbolic and semantic representations),
3. Deep Learning for NLP: summarization, translation, named entity recognition,question answering, document classification, etc.
4. NLP for Deep Learning: parsing (part-of-speecb tagging), tokenization, sentence detection, dependency parsing, semantic role labeling, semantic dependency parsing, etc.
The second edition of this workshop aims at demonstrating recent and future advancesin semantic rich deep learning by using Semantic Web and NLP techniques which canreduce the semantic gap between the data, applications, machine learning process, inorder to obtain a semantic-aware approaches. In addition, the goal of this workshop isto bring together an area for experts from industry, science and academia to exchangeideas and discuss results of on-going research in natural language processing, structured knowledge and deep learning approaches.

Website: https://sites.google.com/view/deepontonlp-eswc2021/committees

Video: https://drive.google.com/file/d/1Zi9hiIykpHejjK_MG8AkWaRn4avK4jAy/view?usp=sharing

Organisers: Pasquale Lisena and Albert Meroño-Peñuela

The success of Semantic Web technology has boosted the publication of Knowledge Graphs, and several technologies to access them have become available covering different spots in the spectrum of expressivity: from the highly expressive SPARQL to the controlled access of Linked Data APIs, with GraphQL in between. Many of these technologies have reached industry-grade maturity. Finding the trade-offs between them is often difficult in the daily work of developers, interested in quick API deployment and easy data ingestion. In this tutorial, we will cover this in-between technology space, with the main goal of providing strategies and tools for publishing Web APIs that ensure the easy consumption of data coming from SPARQL endpoints. Together with an overview of state-of-art technologies, the tutorial focuses on two novel technologies: SPARQL Transformer, which allows to get a more compact JSON structure for SPARQL results, decreasing the effort required by developers in interfacing JavaScript and Python applications; and grlc, an automatic way of building APIs on top of SPARQL endpoints by sharing queries on collaborative platforms. Moreover, we will present recent developments to combine the two, offering a complete resource for developers and researchers. Hands-on sessions will be proposed to internalize those concepts with practice.

Website: https://d2klab.github.io/swapi-eswc21/

Video: https://www.youtube.com/watch?v=pR2eOuVib-4&feature=youtu.be

Organisers: Raúl García-Castro, John Davies, Grigoris Antoniou and Carolina Fortuna

The concept of digital twins, as virtual replicas of physical entities, has gained significant traction in recent years in a range of domains such as industry, construction, energy, health or transport. Digital Twins can be used to view the status of the twinned physical object, without the need to interrogate the object itself. The digital twin can be queried by other software without the need to query the device itself thus relieving pressure on devices, which typically have very limited computational capabilities. Digital twins can also be used for monitoring and diagnostics to optimize device performance without impacting on the physical device.
Digital twins require unambiguous descriptions of both the entity and its digital counterpart, as well as the ability to integrate data from heterogeneous sources of information (including real-time data) and to interact with the physical world. Given these requirements, semantic technologies can play a significant role in the real-world deployment of digital twin technology.
The aims of the SeDIT workshop are twofold. Firstly, to drive the discussion about current trends and future challenges of semantic digital twins. Secondly, to support communication and collaboration with the goal of aligning the various efforts within the community and accelerating innovation in the associated fields.

Website: https://sedit.linkeddata.es/

Video: https://youtu.be/KJz3e3bSvY0

Organisers: David Chaves-Fraga, Anastasia Dimou, Pieter Heyvaert, Freddy Priyatna and Juan F. Sequeda

More and more Knowledge Graphs (KGs) are generated for private use, e.g., Siri, Alexa, or public use, e.g. DBpedia, Wikidata. While techniques to automatically generate KGs from existing Web objects exist (e.g., scraping Web tables), there is still room for improvement. Initially, generating KGs from existing datasets was considered an engineering task with some ad-hoc approaches, however, scientific methods with more declarative-oriented techniques have recently emerged. In particular, several mapping languages for describing rules to generate knowledge graphs and processors to execute those rules emerged. Addressing the challenges related to KG construction requires both the investigation of theoretical concepts and the development of tools and methods for their evaluation. KGC is a full-day workshop with a special focus on knowledge graph construction methods that involve or analyze the roles of the users in these processes. The workshop includes a keynote and a panel, as well as (research, in-use, experience, position, tools) paper presentations, demo jam and break-out discussions. Our goal is to provide a venue for scientific discourse, systematic analysis and rigorous evaluation of languages, techniques and tools, as well as practical and applied experiences and lessons-learned for generating knowledge graphs from academia and industry. The workshop complements and aligns with the activities of the W3C CG on KG construction.

Website: https://kg-construct.github.io/workshop/2021/

Video: https://youtu.be/foBdgpVHW-U

Organisers: Dennis Diefenbach, Andreas Both and Pierre Maret

Knowledge Graphs are designed to be easily consumed by machines, but they are not easily accessible by end-users. Question Answering (QA) over Knowledge Graphs (KGs) is seen as a technology able to bridge the gap between end-users and Knowledge Graphs. In the last years a lot of research was carried out to solve the problem of QA over KGs, but constructing a QA system over a new KG for non-expert users is still cumbersome and time-consuming. The aim of this tutorial is to address this issue and enable researchers as well as practitioners to build QA systems and measure their quality. We will show how recently developed technologies, like QAnswer and Qanary, allow constructing, customizing, evaluating, and optimizing QA systems on RDF datasets using a lightweight approach.

Website: https://qanswer.github.io/QA-ESWC2021/

Video: https://drive.google.com/file/d/1obxZxxoL46Azt04Guum78TPmJIstRXv3/view

Organisers: Davide Buscaldi, Danilo Dessì, Mauro Dragoni, Diego Reforgiato Recupero and Harald Sack

People use online social platforms to express opinions about others, products, and/or services in a wide range of domains, influencing the point of view and behavior of their peers. Understanding individuals’ opinion and satisfaction is a key element for businesses, policy makers, organizations, and social institutions to make good decisions. Sentiment Analysis methodologies have been investigated and employed by researchers to mine sentiment and emotions of people and provide methodologies and resources to all the stakeholders. In particular, sub-symbolic representation and learning have been deeply explored. However, how to augment with explicit semantics and explain what sub-symbolic models learn are questions that still remain open and needs further research and development. Ontologies and Knowledge Graphs can be can drive the analysis, understanding, and traceability of models for an enhanced and human-comprehensible Sentiment Analysis.

Website: https://danilo-dessi.github.io/xsentiment/

Video: https://drive.google.com/file/d/1Vfgj66HiAo4vuLkn8yrSomTZkPLwTNa8/view?usp=sharing

Organisers: Silvia Chiacchiera, Joana Francisco Morgado, Gerhard Goldbeck and Martin Thomas Horsch

Semantic technologies are nowadays widespread as a component of Industry 4.0 architectures, but their uptake is not at the same level in all fields of industrial engineering; fragmentation and duplication of efforts remain an issue. In DORIC-MM we welcome both ontologists and experts from all domains of materials and manufacturing that are willing to engage in a genuine discussion from which both sides will benefit. We ask ontologists (“providers”) to show concrete and convincing examples of how semantic technologies can be used in this field, also welcoming negative results and lessons learned. On the other hand, we encourage domain experts (“users”) to bring to the table their current challenges on data documentation, integration and management. In both cases, a second party will provide constructive feedback.
DORIC-MM is co-funded by the OntoCommons project. It will draw a map of the semantic landscape in the context of materials and manufacturing. In particular, its outcome will consist of populating a registry of semantic assets created within the project and writing a report document that critically captures the status of the field to advance it.

Website: https://ontocommons.eu/doric-mm-2021

Video: https://www.youtube.com/watch?v=Bd2LL41yVeg