Tag: XDP

Posted on

Karl Hammar defended his PhD thesis on Ontology Design Patterns

On September 29, Karl Hammar successfully defended his PhD thesis entitled “Content Ontology Design Patterns: Qualities, Methods, and Tools”. The thesis was first presented by the opponent, professor Harald Sack, FIZ, Karlsruhe, Germany, who then continued to discuss the research results and future ideas with Karl. The three members of the examining committee then continued the discussion with Karl, before unanimously deciding to approve his thesis and award the PhD degree to Karl. At which point we could all congratulate him to a an interesting defence and an excellent thesis. In particular, his three supervisors, Henrik Eriksson (LiU), Eva Blomqvist (LiU) and Vladimir Tarassov (JTH) were of course the first to congratulate him, followed by colleagues and family.

Karl Hammar’s research has aimed to combine quantitative and qualitative research methods, primarily based on five ontology engineering projects involving inexperienced ontologists, studying how Ontology Design Patterns (ODPs) can support that specific group of users. A series of ontology engineering workshops and surveys provided data about developer preferences regarding ODP features and quality, ODP usage methodology, and ODP tooling needs. Other data sources were ontologies and ODPs published on the web, which have been studied in detail. To evaluate tooling improvements, experimental approaches provided data from comparison of new tools and techniques against established alternatives.

The analysis of the gathered data resulted in a set of measurable quality indicators that cover aspects of ODP documentation, formal representation or axiomatisation, and usage by ontologists. These indicators highlight quality trade-offs: for instance, between ODP Learnability and Reusability, or between Functional Suitability and Performance Efficiency. These are things that ontology engineers need to keep in mind when using ODPs in their ontologies, and in particular if they are inexperienced ontologists. Furthermore, the results demonstrated a need for ODP tools that support three novel property specialisation strategies, and highlighted the preference of inexperienced developers for template-based ODP instantiation, neither of which were supported in prior tooling. The studies also resulted in improvements to ODP search engines based on ODP-specific attributes. Finally, the analysis showed that a specific ontology engineering methodology, the eXtreme Design (XD), should include guidance for developer roles and responsibilities in ontology engineering projects, suggestions on how to reuse existing ontology resources, and approaches for adapting XD to project-specific contexts. Karl therefore proposed a new version of the XD methodology, specifically covering these aspects.

The thesis can be found here.



 

 

 

 

 

 

 

Congratulations Karl!

Posted on

LiU Semantic Web group at ESWC2017

This week a couple of us have been at ESWC2017 in Portoroz, Slovenia. Eva Blomqvist was the general chair of the conference this year, hence, this was the culmination of a whole year of hard work for her. Olaf Hartig is the proceedings chair (proceedings part 1 and 2). He could not attend the conference this year, but has done great job with the Springer proceedings, and the upcoming post-proceedings volume with poster and demo papers among other things. In addition to this, Karl Hammar, was one of the organisers of the Modular Ontology Modeling with Ontology Design Patterns tutorial, together with Pascal Hitzler, Adila A. Krisnadhi, Agnieszka Lawrynowicz and Monika Solanki. In particular, Karl ran the hands-on session with his tool for ODP-based modelling in WebProtégé (called XDP). Finally, Henrik Eriksson, presented our EU-funded project VALCRI in the project networking session, and in the poster session.

The overall conference was interesting as always, and included a lot of networking opportunities, as well as interesting work to take a closer look at. A quick summary of some of the major events:

Crosbie

Kevin Crosbie, from Ravenpack, the first keynote speaker talking about how to model events in order to use them for predicting financial markets. Very interesting talk, describing how Ravenpack work with their data products and apply technologies very similar to Semantic Web, although technically not using the W3C standards, such as RDF.

Panel

At the end of the first day, Aldo Gangemi chaired a panel about the future of academic publishing, discussing the challenges and opportunities that lie ahead. It is clear that something needs to be done about both the reviewing situation in our field, the open access issue, and we want more focus on “eating our own dog food”. The discussions were also related to the paper that later won the best student paper award, on Linked Data Notifications.

Sheridan

Second keynote speaker, John Sheridan, from the National Archives in the UK, described how the National Archives heavily rely on Semantic Web technologies and standards to solve their archiving tasks. However there are also challenges of course, which can hopefully be solved by working together: academia and society at large. Particularly interesting for us at LiU to hear that the National Archives is in great need of a better solution for modelling trust and uncertainty in their data, which could be a potential use case for the recent research results on RDF* and SPARQL* by Olaf Hartig.

Dinner

Nice conference dinner at the beach, and a chance for the general chair to thank all the people in the organising committee.

Poster

Poster session with lots of interesting interactions and discussion, here with Diego Reforgirato, who later won both the best poster and best demo awards.

Unfortunately, we did not take any picture of the last keynote, Lora Aroyo, who gave a very interesting keynote on the last day. She started with an overview of the evolution of the field, pointing out that studying and using people to acquire knowledge has always been a central part of our research. However, by over simplifying, and trying to fit every answer into yes/no categories, we can introduce wrong conclusions. She means that we need to be aware of ambiguity and diversity in opinions, that there is usually not one true answer, and instead turn that to our advantage. Lora showed a vector-based model to represent diversity in opinions.

Finally, Aldo Gangemi will be the next general chair of ESWC in 2018, and he made a series of interesting promises for the next year, among others: double-open review process, improvements in the online pre-prints of the proceedings and the dataset, a resources track and an industry session á la ISWC, and better music in the social events. We all wish him the best of luck with the next conference, and we are excited to see all the innovations next year!