| layout | title | permalink |
|---|---|---|
page |
Fusepool P3 Development |
/ |
The goal of Fusepool P3 project is to make publishing and processing of public data as linked data easy. For this purpose Fusepool P3 develops a set of of software components that integrate seamlessly by well defined API basing on Linked Data Best Practices and the Linked Data Platform standard.
To learn more about the overall platform as well as the APIs we propose see: http://fusepool.gitbooks.io/the_fusepool_p3_platform/content/d51-deliverable.html
To learn how to write a transfomer in Java: https://github.com/fusepoolP3/p3-transformer-howto/blob/master/transformer-howto.md
The recommended way to try it out is to run the docker image. If you've installed Docker you can start the Fusepool reference implementation by executing:
git clone https://github.com/fusepoolP3/p3-platform-reference-implementation.git p3
cd p3/marmotta/
docker-compose up
The above will start an instanced backed by the Apache Marmotta LDP implementation, see virtuoso for using Virtuoso Open Source.
We also run an instance of the reference implementation on: http://sandbox.fusepool.info/
The following are applications providing implementation or basing on the Fusepool APIs.
The transforming containers API is implemented by the P3 Proxy which can be backed by any compliant LDP implementation. A Marmotta backed instance is available at http://sandbox.fusepool.info:8181/.
The Transforer containers API is implemented by a growing number of services that allow transforming data. Check out the following transformers:
The dictionary matcher provides transformers the recognize entities from a SKOS taxonomy. For example the transformer with URI http://sandbox.fusepool.info:8192/?taxonomy=http://data.nytimes.com/descriptors.rdf will find mentions of New York Times category in a textual content.
To try it out witch cURL:
curl -X POST -d "Frauds and Swindlings cause significant concerns with regards to Ethics." "http://sandbox.fusepool.info:8301/?taxonomy=http://data.nytimes.com/descriptors.rdf"
The sources and more information about this transformer are available here: https://github.com/fusepoolP3/p3-dictionary-matcher-transformer
The Batch Refine Transformers uses an Open Refine configuration file to transform some input data according to the OpenRefine transformation rule. For example this can be used to generate clean RDF.
The Geo Enriching Transformers enriches RDF data containing geographical locations with points of interests around these locations. The locations are taken from an URI that can be specified as a query parameter in the URI of the transformer. For example the transformer with URI http://sandbox.fusepool.info:8193/?data=https://raw.githubusercontent.com/fusepoolP3/p3-geo-enriching-transformer/master/src/test/resources/eu/fusepool/p3/geo/enriching/test/farmacie-trentino-grounded.ttl will enrich data with nearby pharmacies (assuming the data describes locations close to a pharmacy of Trentino).
The sources and more information about this transformer are available here: https://github.com/fusepoolP3/p3-geo-enriching-transformer
The Pipeline Tranformer is a tranformer executing a list of (other) transformers in sequence.
The Fusepool project partner OpenLink Software provides serveral transformers to transform data to RDF.
OpenLink Software also provides serveral transformers automatically generating annotations to textual content.
It is a graphical user interface to list the available transformers and provide functionality for creating pipelines.
This demo app provides a simple interface to check pubs, restaurants, pharmacies, accomodations, museums and - optionally - events in Tuscany and Trentino Region on a map. You can click the location you are interested in to see what can be found nearby. To use event data from a specific resource, use the "events" attribute in the URL in which provide the URI of the resource:
Need help or have suggestions? Get in touch with the developers with the mailing list at [email protected] or raise issues in github.
Fusepool P3 is partially funded by the 7th Framework Program for Innovation, under grant 609696. For further details about the project, please refer to the official webpage.