EKAW 2014

From www.b-kaempgen.de
Revision as of 12:00, 12 December 2014 by Benedikt Kämpgen (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search


Short info about the trip

  • Venue title: 19th International Conference on Knowledge Engineering and Knowledge Management (EKAW) 2014
  • Related project: Software-Campus (LD-Cubes, LisKo)
  • Venue location: Linköping, Sweden
  • Venue date: 2014-11-26
  • Website: http://www.ida.liu.se/conferences/EKAW14/home.html
  • Short description: Presented research paper on "Building the Global Cube". Around 60 participants (one track venue). Although diversity was topic, few papers discussed data integration problems. Topics: ontology design patterns, modelling (e.g., of medical guidelines and change of meaning of hashtags).
  • Outcome: Good feedback to presentation (alignment of statistics from different organisations seen as important but still mainly manual task). One great keynote about ontology engineering by Oscar Corcho. One design pattern-focussing keynode by Pascal Hitzler. One very abstract keynote by a philosopher (Arianna Betti).
  • Other participants: Pascal Hitzler, Oscar Corcho, Isabelle Augenstein, Frank van Harmelen, Vinay Chaudhri
  • Good experience: Great social events: Ice Hockey, Visualisation center Norrköping, 3D virtual observatory. Poster, "LSD Dimensions" by Albert Meroño-Peñuela, about data cubes found on the web (3,000 dimensions). Martina Freiberg from Würzburg on Spreadsheet-based Knowledge Acquisition for physicians.
  • Bad experience: Some papers not as interesting as titles suggested.

Longer description

Conference about knowledge management. Topics mainly: modelling, logics, Semantic Web concepts.

Keynotes:

  • Pascal Hitzler: Rather position presentation. Highlighting the importance of ontology design patterns. Linked Data vocabularies do not help. FOAF in the end only is a collection of URIs. Everyone is using it with a different semantic in mind. Similar for SKOS. With ontology design patterns and DL we can constrain meaning. Unclear: How to use design patterns in practice.
  • Oscar Corcho: 10 origins of ontology (KA provides one, number of citations unknown); Brave new world. Alphas, Betas, Gammas, Deltas oder Epsilons. 1932 erschienener dystopischer Roman von Aldous Huxley. Alphas (>60, few, Upper-level ontologies, Dissertation on one single term such as "Action"), Betas (Oscar, >40, more, domain specific ontologies), Gammas (Linked Data, vocabularies), Deltas (data publishers), Epsilons (Web developers with schema.org). Then, there are the savages. How to address these? Competency questions = user stories; Deadlocks (1. no reuse at the beginning, 2. inferences 3. textual descriptions, 4. use simple tools, Excel). Problem: No clear communication of what to aim for with ontologies (discussion!).
  • Arianna Betti (Concepts in Motion). Alpha. Her work is about "the Model of Science". Very abstract "semantic core" vs. "semantic margin". Savage. Looked into one book, for the development of one term. Such people really need to be convinced.

Feedback to my own presentation [1]

  • Generally well received. Several people approached me later; but mainly from a management perspective of such correspondences.
  • Oscar liked idea about the global cube.
  • Oscar did not like my last sentence saying that Wolfram Alpha and Eurostat should publish the same number of GDP per Capita.
  • Frank van Harmelen was interested about the manual effort in creating the correspondences. He wondered whether parts can be automated and whether such information can be retrieved somewhere. Also he was wondering about possible user interfaces of an integration and analysis system.

Awards

  • Adaptive Knowledge Propagation in Web Ontologies Pasquale Minervini, Claudia D'Amato, Nicola Fanizzi and Floriana Esposito - link prediction, e.g., predicting the affiliations of AIFB staff members to research groups with AIFB Portal dataset. statistical learning.

This event was attended by: Benedikt Kämpgen