Web Engineering (WE)

Main research interest

In the early 1990’s the world of "information" shifted from so-called administrativedata to textualand later multi-mediadata. For roughly a decade the new information world became a Web of documents. The Web Engineering group was one of the pioneering groups studying how to do automatic personalization (aka adaptation) on websites in many different application areas.

In the most recent decade (and even before that) the research interest is shifting from a Web of text (and other media) to a Web of semantic data, in which bits of data are linked together, using RDF as its main language, and URIs (and URLs) for identifying and locating these data. This has led to huge distributed (and unreliably linked) data stores for which theissue of querying and retrieval becomes a seriouschallenge. The Web Engineering group has achieved fundamental groundbreaking results in the design and development of the technology for handling (indexing, searching,retrieving) this huge amount of graph-structured data.

Success stories

Already in 1993 the first open on-line course (on the topic of hypermedia) was launched and delivered through the then still new Web. In 1996 this open course was made adaptive, or personalized based on the learner’s browsing behavior. Where a successful MOOC today would hope to reach around 0.0001% of the Internet population the open on-line hypermedia course easily reached 0.001% (i.e. 10 times more).

The work on our generic adaptation platform, initially called AHA! and later GALE lead to a number of world’s firsts: an adaptive research paper, an adaptive conference presentation, even an adaptive keynote talk (all around 2006) and the first adaptive PhD thesis (2012). The research field on user modeling and adaptation is being "served" by User Modeling Inc. (that organizes the yearly ACM UMAP conference) and prof. De Bra is president of this organization.

In the data engineering area, the group has developed one of the first structural-index-based triple stores and have their results on indexing and querying in graph data regularly appearing in flagship data management conferences and journals. Their investigations are in close collaboration with leading and cutting-edge database companies such as Oracle and Neo Technologies. Further impact is also made through scientific membership in the LDBC Graph Query Language Standardization task force, an international academia-industry collaboration.

Project examples

  • AHA! (Adaptive Hypermedia for All) NLnet Foundation. This early funding marked the start of the development of the later AHA! and GALE platforms that have been used for research on adaptation all over the world.
  • CHIP NWO CATCH. This Cultural Heritage project demonstrated how access to cultural information and visits to museums can be personalized using semantic metadata.
  • GRAPPLE EU FP7. This (5M€) project with 15 partners showed how user modeling and personalized learning can be extended across multiple organizations to enable continuous personalization in life-long learning.
  • SeeQR NWO. This project is the first to study engineering of sophisticated structural-indexing and query processing solutions for massive RDF and graph data, on solid theoretical foundations.