Technical details

AgriVIVO works as a common store of URIs and relations to interlink the data managed in the existing communities and databases. Data management can remain decentralized as well as data browsing, as VIVO’s search functionalities can be integrated in other websites.

AgriVIVO integrates key existing sources of information about agricultural researchers by discipline, organization, or topic, using a consistent ontology framework.

It builds upon the  essential features of VIVO:

  • the ontology approach to model the organization of agricultural research and the relations between all its entities;
  • the RDF framework to build and update the ontology and to store URIs for all the entities identified in the ontology and the relations between them, expressed through properties from the most appropriate RDF vocabularies and from specialized AgriVIVO vocabularies;
  • data harvesting tools for parsing, mapping and importing data from systems that use different machine-readable mechanisms to expose their data;
  • special algorithms for disambiguating entities imported from different sources, also exploiting what has already been done by other initiatives in this area;
  • internal and external authority data, so that AgriVIVO can offer a reference environment for any other information system: alongside existing authority data for geographic information and for agricultural concepts, AgriVIVO will integrate and align internal and external authority data for institutions and for people.

AgriVIVO customizations

Ontology

Ontology extensions can be implemented in any VIVO installation to add greater specificity, but the core VIVO reasoner will always be able to infer more general type statements and thanks to the sharing of the core ontology consistent queries can be run across Institutions, even if some Institutions use local ontology extensions.

The VIVO concept and RDF model can be adapted to fit the way agricultural research works and implement a "research discovery tool" in agriculture by extending the core ontology and “hiding” classes and properties that relate to entities that don’t belong to the “real world” of agricultural research.

The AgriVIVO model shares with the VIVO model the basic entities of expert profiles, Organizations, projects, publications, events, geographic locations and all reciprocal relations. Most of the work on the ontology extension consists in the adjustment of sub-classes for these entities: in the VIVO model, specific “types” of the above entities are sub-classes, so for instance sub-classes have to be created under the Organization class to include non-academic organization types.

See the AgriVIVO vocabulary extension and an overview of the AgriVIVO core ontology (including VIVO classes that are core also to AgriVIVO)..


Search portal

The data are harvested to the main AgriVIVO store (a customized VIVO installation extended with the AgriVIVO vocabulary and several custom importers).

However, the search portal has been built separately as a user-friendly interface to search and browse the data.

The search portal imports data from the AgriVIVO triple store and indexes them with Solr. It offers a Solr-based faceted browsing and an advanced search, plus other maps and charts to access the data.

The search portal is also a demonstration of how the data in the AgriVIVO triple store can be re-used by other applications. 

More information on how to re-use AgriVIVO data in other applications will be available soon.