CIWS researchers stand out in the most important Semantic Web Conference

"A Formal Framework for Comparing Linked Data Fragments", "IMGpedia: a Linked Dataset with Content-based Analysis of Wikimedia Images", and "Answering Visuo-semantic Queries with IMGpedia", were the articles of the DCC that obtained "Best paper", "Best Student Resource Paper" and "Best Poster" respectively at the International Semantic Web Conference (ISWC), the most important conference in the area of ​​the Semantic Web, which took place this year between October 21 and 25 in Vienna.

"A Formal Framework for Comparing Linked Data Fragments", written by DCC professor Jorge Pérez, together with Ian Letter, DIM student and Olaf Hartig, researcher at Linköping University in Sweden, is an article that formalizes a new way of access linked data on the Web called "Linked Data Fragments" (Linked Data Fragments).

"The idea is that different data servers on the Web expose interfaces with different capabilities, and until now there was not a good way to compare these capabilities. For example, if a user or application wanted to complete a certain task with data from the Web, there were no formal tools to help users or applications decide which interfaces to use, "explained Professor Pérez.

The academic highlighted that the Research Center of the Semantic Web (CIWS), made up of DCC-UChile and DCC-PUC academics, has established itself as one of the most important centers in the world on the subject.

"The awards received this year at the most important international conference on Semantic Web research, add to several awards that researchers have previously received together (in ESWC 2005, ISWC 2006, ESWC 2007, WWW 2012, and ISWC 2016) , even in 2016, one of the CIWS jobs was awarded as the most impactful work in the area written 10 years ago. This is a recognition of the efforts of several academics who have positioned CIWS well above in the area of ​​Data and Semantics of the Web. "

On the other hand, the articles "IMGpedia: a Linked Dataset with Content-based Analysis of Wikimedia Images", and "Answering Visuo-semantic Queries with IMGpedia", written by Sebastián Ferrada, Aidan Hogan and Benjamín Bustos, are part of the thesis of Master's degree, carried out by the doctoral student Sebastián Ferrada.

IMGpedia is a linked database that provides similarity relationships between Wikimedia images and links to other knowledge bases, to obtain data about the context of the image. "The beauty of this is that you can make semantic queries that involve images, such as obtaining sixteenth-century paintings that are on display in the Louvre or given a set of images of European cathedrals, obtaining similar images that are museums", explained our doctoral student Sebastián Ferrada and added: "IMGpedia is a relevant advance, since it proposes an original / novel way to deal with the multimedia content of the Web, and go beyond annotations made by people and analyze the content of the image" he explained.
 
Meanwhile, Professor Aidan Hogan pointed out that these jobs are just a beginning and that there is much to improve. "We are only in step one with an initial prototype. Sebastián is going to continue working in IMGpedia during his doctorate, since the results are very encouraging ". He added: "These awards have high visibility, since ISWC is the most important conference in the Semantic Web area. Sponsors of the conference include Google, Fujitsu, Siemens, Oracle, IBM, etc. There are papers from universities such as Oxford, MIT, Stanford, etc. In fact, the "Best Student Resource Paper 2017" award was shared with Stanford researchers, which is certainly impressive, "he said.

Finally, Professor Benjamín Bustos stressed that having received this recognition in the ISWC 2017 is very relevant, "since it shows that through the collaboration of professors from different areas in computing, and together with the work of our graduate students can be obtained results that have a high impact for the scientific community internationally, "he concluded.

------

Comunicaciones DCC