START Conference Manager    

Using ElasticSearch as an INSPIRE metadata backend

Guillaume SUEUR

(Submission #173)


Abstract

Since the early ages of INSPIRE discovery services we are using tools to store and retrieve metadata. These tools were designed more than ten years ago now, and maybe a new set of tools can completely change the way we deal with metadata. What we propose here is a presentation of an implementation around ElasticSearch for dealing with metadata records as well as data records themselves. Build on top of ElasticSearch, a comprehensive query API is able to search, filter, extract these different records from the ES database itself. The discussion will then focus on a new query paradigm : if a search engine is able to deal with thousands or millions of records, and to retrieve a relevant set of results in a few seconds, is composing, storing and querying metadata is still relevant ? To make a parallel with the web world, we would like to bring the INSPIRE world from the Yahoo paradigm, when people were manually referencing websites, to the Google paradigm, where the search engine is able to collect, organize, promote websites directly from the analysis of their content. Applied to the INSPIRE world, this supposes to completely mix what is currently done with CSW and WFS, and to be able to put in the search engine enough intelligence (dictionaries, semantics...) to make the indexing relevant from different usages point of views.

Categories

Topic Area:  [2.3] Technologies and tools to support implementing, using and assessing the technical implementation of INSPIRE
Abstract Type:  Oral Presentation

Additional Fields

 
Comments:   Metadata, CSW, WFS


START Conference Manager (V2.61.0 - Rev. 4840)