Help Document for GeoLink Demo

Leveraging Semantics and Linked Data for Data Sharing and Discovery

Introduction

Welcome to the GeoLink Demo, a Linked-Data-driven Web portal demonstrating the power of ontologies and Linked Data in supporting data sharing in the domain of Geoscience. GeoLink organizes and links the data from a number of major data repositories funded by NSF, including BCO-DMO, R2R, DataONE, IODP, IEDA, LITER, and MBLWHOI. By performing a single query, a user can obtain results from all participating repositories. GeoLink Demo consists of three views, namely tabular view, graph view, and map view, which allow the user to examine the data from three different perspectives.

Highlights

- Federated query: each query is performed across all participating data repositories, while this complexity is hidden from the end user.
- Entity coreference: information about the same entities across different data repositories is dynamically combined.
- Multiple views: each entity can be viewed from the perspectives of a table, a graph, and a map (if coordinates are available).

Main Data Access Board

When first entering the demo at http://demo.geolink.org, you will be presented with the interface on the right which provide access to the major entities hosted by the participating data repositories. These entities include Datasets, Cruises, Vessels, Instruments, Researchers, and so forth. You can examine these entities by clicking the different viewer icons under the entity type icon: for a tabular view, for a graph view, and for a map view (if the type of entity has geographic coordinates).

Tabular View

Tabular View shows the information about entities in the form of a table. For example, if you go to the "Researchers" panel, type in "Wiebe", and click the search button, the system will find all researchers whose name contains "Wiebe". If you click "Wiebe, Peter", you will be able to see related information about this researcher, such as his affiliation, cruises led by him, and so forth. A cool feature of the Tabular View is that it allows the follow-your-nose style data browsing which enables users to browse the data by simply clicking the related links. For example, after seeing the information about "Wiebe, Peter", you can continue clicking the affiliation of this researcher to learn information about this institute.

Graph View

Graph View allows users to examine the entities and their relations from a perspective of nodes and links. Graph is the native form that Linked Data have been stored, and a graph view intuitively shows the links between entities. By clicking the from the Tabular View, you can see the information about "Wiebe, Peter" from as a graph (Note: you can also access the graph view directly from the main access board). By right clicking the nodes, you can learn more information about each entity by choosing the "Learn More..." from the context menu. Graph View also allows users to find relations between entities. By going to "Relationship Finder" on the second tab of the right panel, one can type in two entities (or one entity and an entity type) and discover the relations between them. For example, one can find all the cruises led by "Wiebe, Peter".

Map View

Map View provides visualization for data entities that contain geographic coordinates. For example, you can access the map view of "Cruises", click the "search" button, and click the "Map Result" button, to see the cruises on a map. On the left panel, you can see the map layer as you would typically see in a geographic information system. You can check the visibility box to control whether a layer should be displayed on a map. The icon controls whether you want to cluster the results when points are presented on the map. You can also directly click the geographic entities on the map to see detailed information about them.