dr. Willem van Hage

E-mail 
Social networks 

Netherlands eScience Center
Netherlands eScience Center
Science Park 140,
1098 XG Amsterdam
T 06-15886039

Vrije Universiteit Address
Web & Media Group < Computer Science
< Faculty of Sciences
< VU University Amsterdam
De Boelelaan 1081A,
1081 HV Amsterdam,
Room U332
T 020-5988785,
F 020-5987653

I am Technical Lead data management & analytics at the Netherlands eScience Center and guest researcher in the Web & Media group at the VU. I coordinate scientific software development and research projects in various domains (geography, climatology, law, physics, etc.) that have to do with analytics. I do research on information integration, visualization, and event modeling, focussing on improving the speed, transparency, and reproducibility of the research process, with a special interest in explorative and creative research.

My PhD thesis about ontology alignment is available here: “Evaluating Ontology-Alignment Techniques” and my MSc thesis about the border between ontology engineering and information retrieval is available here: “Living on the Edge”.

Research Linked Open Piracy, Space package, SEM

My research mainly deals with Information Integration, Visualization, and Reasoning.

Computer science topics that occupy me are: Ontology Alignment, Integrating Spatial reasoning and Semantics, Trust on the Web, Data Mining, specifically mining knowledge from the World Wide Web, The Semantic Web, Description Logic, combining Logical and Statistical reasoning, Information Retrieval & Extraction.

Some other topics that interest me are: Information Theory, Program Algebra, Logical & Functional programming languages.

Program Committees

I am co-chair of DeRiVE 2012, and DeRiVE 2011.

I am a PC member of ISWC 2013 In Use track, Cooking with Computers (CwC) 2013 at IJCAI, ESWC 2013, ISWC 2012 Evaluation track and In Use track, GiBDA 2012, AIW 2012, ISWC 2011, Web-KR 2011, LSTD 2010. ISWC 2009, ISWC 2009 Doctoral Consortium, and ESWC 2009 Posters & Demos.

I have reviewed papers for the Artificial Intelligence journal the Journal of Web Semantics, the Semantic Web Journal and Information Systems; and for the following conferences, workshops, and summer schools: EKAW 2012, IJCAI 2011, KEOD 2010, FOIS 2010, KCAP 2005-2009, ESWC 2009, ICSD 2009, ECAI 2009, SemEval 2009 workshop (at ACM 2009), ESSLLI 2006, and ASWC 2006.

SPARQL client for R

I wrote a SPARQL client for the R statistical programming language together with Tomi Kauppinen and a number of additional contributors.

Tutorials can be found at:

• http://semanticweb.cs.vu.nl/R/
• http://linkedscience.org/

This client allows you to directly import results of SPARQL SELECT queries into the statistical environment of R as a data frame. That means you can directly perform statistical analysis on data sets on the web. For example, you can use the following R code to get data from the Linked Open Piracy SPARQL end point described below.

library(SPARQL)
endpoint <- "http://semanticweb.cs.vu.nl/lop/sparql/"
q <-
  "SELECT *
   WHERE {
     ?event sem:hasPlace ?place .
     ?place eez:inPiracyRegion ?region .
   }"
prefix <- c("lop","http://semanticweb.cs.vu.nl/poseidon/ns/instances/",
            "eez","http://semanticweb.cs.vu.nl/poseidon/ns/eez/")
res <- SPARQL(endpoint,q,ns=prefix)$results
pie(sort(table(res$region)),col=rainbow(12))

This yields the following overview of the world-wide distribution of piracy attacks:

With a little more effort you can plot correlations.

q <-
  "SELECT *
   WHERE {
     ?event sem:eventType ?event_type .
     ?event sem:hasPlace ?place .
     ?place eez:inPiracyRegion ?region .
   }"
res <- SPARQL(endpoint,q,ns=prefix)$results
restable <- table(res$event_type,res$region)
par(mar=c(4,10,1,1))
barplot(restable,col=rainbow(10),horiz=T,las=1,cex.names=0.8)
legend("topright",rownames(restable),
       cex=0.8,bty="n",fill=rainbow(10))

This yields the following overview of the distribution of kinds of events per region:

Using the ggmap package you can draw an overview map of the same events. The ggmap package requires you to install the mapproj package.

library(ggmap)
q <-
  "SELECT *
   WHERE {
     ?event sem:eventType ?event_type .
     ?event sem:hasPlace ?place .
     ?place wgs84:lat ?lat .
     ?place wgs84:long ?long .
   }"
res <- SPARQL(endpoint,q,ns=prefix)$results
qmap('Gulf of Aden', zoom=2, legend='bottomright') +
  geom_point(aes(x=long, y=lat, colour=event_type), data=res) +
  scale_color_manual(values = rainbow(10))

This yields the following map:

Linked Open Piracy

Together with Véronique Malaisé, Marieke van Erp, I released Linked Open Piracy (LOP), a Linked Open Data version of the ICC-CCS International Maritime Bureau reports of piracy attacks world wide. This data set is linked to Princeton WordNet 2.0 and 3.0, Freebase, DBpedia, Geonames and Exclusive Economic Zone data from VLIZ.


Pirate attacks in the Gulf of Aden from the Linked Open Piracy data set.


An example pirate attack modeled in the Simple Event Model (see below).

the SWI-Prolog space package

Together with Jan Wielemaker I am working on a Prolog library for spatial indexing called the space package. It provides fast nearest neighbor, intersection, and containment queries. The current version can deal with Points, LineStrings, and Polygons as query shapes or as data shapes. Results are “returned” one by one, non-deterministically. This makes writing combined spatial/semantic/temporal query optimizers in Prolog relatively easy. The space package also provides back and forth conversion between a number of geospatial/geosemantic standards, like: GML, KML, WKT, and RDF with GeoRSS Simple and GeoRSS GML or with WGS84 properties.

You can find more about the space package in the workshop article The Space package: Tight Integration Between Space and Semantics. The source code of the space package can be downloaded from or browsed on the GIT repository. Documentation can be found in the SWI-Prolog documentation server.

the Simple Event Model

Together with Véronique Malaisé I designed the Simple Event Model (SEM), an ontology to describe events in terms of “who did what with what to whom, where and when?”, respectively modeled as Actors, Events, Objects, Roles, and Places, which are all Time-stamped. Mappings from SEM to other models are included in the SEM RDF file.


The Simple Event Model (SEM)

Publications Event and Behavior Modeling, Augmented Reality, Ontology Alignment

Teaching Information Retrieval, Ontology Engineering, Web Science minor, MSc projects

I gave the following courses at the VU University Amsterdam at the Faculty of Sciences:

• Information Retrieval for BSc students in Computer Science and IMM. This course is about topics such as crawling, indexing, various term weighting schemes, and evaluation. Lab exercises involve distributed search on a Hadoop cluster with Katta and Pig Latin on Common Crawl and Wikipedia.
• Knowledge and Media for MSc students in Informatics, Information Sciences, and Artificial Intelligence. This course deals with modeling decisions in knowledge intensive applications that involve multimedia.
• Ontology Engineering for MSc students in Informatics, Information Sciences, and Artificial Intelligence. This course treats topics like modeling decisions, ontological commitment, and design patterns in RDF(S), OWL, UML.
• Computer Science Seminar for MSc students in Computer Science. This is a course about literature study in computer science. It teaches basic scientific skills for critical reading, analyzing, summarizing, comparing, presenting, and debating about scientific articles.

I gave guest lectures for the follwing courses at the VU University Amsterdam at the Faculty of Social Sciences:

• Marketing and persuasive communication for BSc students communication sciences. A lecture dealing with term weighting, PageRank, and Search Engine Optimization.

I gave guest lectures for the following courses at the Technical University Delft:

• Information Retrieval for MSc students in Computer Science. A guest lecture about Latent Semantic Indexing.

I gave guest lectures for the following courses at the Universiteit van Amsterdam:

• Research Methods in AI for BSc students in Artificial Intelligence. A guest lecture about the ideas behind modeling in general and specifically with respect to ontologies, criteria to assess the quality of models, and guidelines for good modeling.
• Intelligent Interactive Systems for MSc students in Information Sciences. A guest lecture giving an overview of localization: from methods for indoor and outdoor geodesy through encoding and indexing to knowledge intensive location-based applications.

I coordinated the design of the Web Science minor at the VU University Amsterdam's Faculty of Sciences.

I gave a short practical business course on Information Extraction from the Web and Semantic Web techniques using SWI-Prolog in 2011.

I supervise the following PhD students: Davide Ceolin, Jesper Hoeksema.

I supervised the following PhD students: Valentina Maccatrozzo, Martine de Vos

I supervised the following MSc students: Tobin Hupsel, Sander Koers, Aäron Jansen, Juan Manuel Coleto Gómez, Vincent Osinga, Bas van Dijk, Thomas Günther, Wiard Vasen, Vikram Ghotgalkar.

I supervised the following BSc students: Niek Siekman, Renske Augustijn.

I was second reader of the theses of the following MSc students: Lieke Verhelst (TUD), Balthasar Schopman, Jevon van Dijk, Andrea Bravo Balado.

Projects and Networks NewsReader, COMMIT Metis, Poseidon, DeRiVE, OAEI

• NewsReader: (project leader together with Piek Vossen) European FP7 STREP project about event-based access to the stories behind financial news and the stock market. The project website can be found at http://www.newsreader-project.eu/.

  ---

• COMBINE 2011-now: (principal investigator) The COMBINE project is an applied research project funded by ONRG. Within the COMBINE project we will set up a framework for the cross-referencing of information from heterogeneous sources (webpages, Semantic Web, spreadsheets, databases, etc.) relying on Web standards. I am the principal investigator of this project.

  ---

• METIS 2011-now: Research in the METIS project is just starting up. My part of this project will mainly be about geospatial reasoning, streaming (RDF) data, provenance and trust management.
  
  Together with Laura Hollink (TU Delft) and François Scharffe I organize the Interlinking New York Times task in the Instance Matching track of the OAEI 2011, a competetive evaluation of ontology alignment between the New York Times subject vocabulary, DBpedia, Freebase, and Geonames.
  
  I wrote a SPARQL client for the R statistical programming language.

  ---

• Poseidon 2008-now: Currently, I do research on geospatial reasoning in the maritime domain within the Poseidon project.
  
  In 2011 Véronique Malaisé, Marieke van Erp, and I set up the Linked Open Piracy (LOP) data set, describing piracy attack events world wide.
  
  Together with Maarten van Someren and Gerben de Vries I organize the Modeling Moving Objects NL 2010 workshop.
  
  Together with Laura Hollink (TU Delft) and Véronique Malaisé I organize the OAEI 2010 very large crosslingual resources (VLCR) task, a competetive evaluation of ontology alignment between the New York Times subject vocabulary, DBpedia, and the GTAA thesaurus of the Netherlands Sound and Vision video archive.
  
  Within Poseidon I work on the SWI-Prolog space package, which provides (geo)spatial indexing for Prolog. This allows fast nearest neighbor, intersection, and containment queries, aswell as conversion between common geospatial standards.
  
  In 2009 Véronique Malaisé and I set up the Simple Event Model (SEM) for describing events in Poseidon and other projects at the Web & Media group, like CHIP and AGORA.
  
  Within the Poseidon project I contribute to the following open source software projects:
  • The XDIG interface to Description Logic servers for SWI-Prolog;
  • YSR, a forward chaining OWL Horst fragment reasoner for SWI-Prolog;
  • spatialindex, a C++ library containing various data structures for spatial indexing;
  • SWI-Prolog's C++ interface; and
  • the LinkedGeoData RDF rendering of OpenStreetMap shapes.

  ---

• Knowledge Web & OAEI 2006 and OAEI 2007: Within the heterogeneity working group (2.2) of Knowledge Web (FP6-507482) I co-organized the Ontology Alignment Evaluation Initiative 2006 and 2007. A full report about this can be found in my PhD thesis.
  
  In 2007 I organized the Food Thesaurus Alignment Task and the Environmental Thesaurus Alignment Task at the OAEI 2007 in cooperation with the Food and Agriculture Organization of the United Nations and the National Agricultural Library of the United States Department of Agriculture.
  More information about the tasks can be found on the Food and Environment Thesaurus Alignment Task 2007 website. (OAEI 2007 food & environment presentation)
  A report can be found in: “Results of the Ontology Alignment Evaluation Initiative 2007, Euzenat, J., A. Isaac, C. Meilicke, P. Shvaiko, H. Stuckenschmidt, O. Šváb, V. Svátek, van Hage, W. R., M. Yatskevich”
  
  In 2006 I organized the Thesaurus Alignment Task at the OAEI 2006.
  More information about the task can be found on the Thesaurus Alignment Task 2006 website. (OAEI 2006 food presentation)
  A report can be found in: “Results of the Ontology Alignment Evaluation Initiative 2006, Euzenat, J., M. Mochol, P. Shvaiko, H. Stuckenschmidt, O. Šváb, V. Svátek, van Hage, W. R., M. Yatskevich”
  

  ---

• Virtual Laboratories for e-Science (VL-e) 2004-2008: Within the Adaptive Information Disclosure (AID, SP2.2) working group of VL-e I worked on web services for ontology mapping and knowledge discovery.

  ---

• TNO projects I participated in 2004-2008: VL-e, Augmented Multi-media Interaction (AMI), Cooperative Vehicle-Infrastructure Systems (CVIS), Operation room of the near future (OKTO).

Martial Arts

Currently I practice shotokan karate (松濤館流 空手) at the SKCA with many teachers amongst which Tom Kompier and Bregje Kaars-Sijpesteijn, where I am a first-degree black belt (shodan, 初段 空手). I am a first-degree black belt jūjutsu (shodan, 初段 柔術) with the JBN. I used to practice at the dojo of the USC in Amsterdam with Edwin Hiense (JBN). I am a green belt judo (third kyū, 三級 柔道) with Eric Bos (JBN) when he still taught at the USC.

me performing a  kata guruma throw
on/with Karien van Assendelft

Bass guitar

I play bass guitar in the (for lack of a better name for the genre) dutch progressive-death-metal-reggae-funk-rock band Gemeengoed.

Computer Generated Art

Below is a gallery of computer generated pictures I made between 1997 and 2001. They were all made using Adobe Photoshop's “Filters” menu and nothing else. (i.e. no paint brushes etc.)

©1997-2001 Willem Robert van Hage

Paradox of Choice

I agree with Barry Schwartz: More choice in our everyday life leads to less happiness.

Free will without non-determinism

I find Daniel Dennett's view on free will is very convincing.