Semantic Enrichment and Search with Linked Open Data: A Case Study on Environmental Science Literature
From June to December 2012, we're running a small exploratory project funded by JISC to experiment with Linked Open Data ontologies for environmental science. The consortium consists of the University of Sheffield, the British Library, and HR Wallingford.
Follow EnviLOD (through this blog) or on Twitter: @GateAcUk #envilod
Aims of the Project
The overall aim of EnviLOD is to demonstrate the value of using Linked Open Data (LOD) vocabularies in the field of environmental science. Firstly, EnviLOD will tackle the problem of LOD domain vocabulary enrichment and interlinking. Tools for efficient LOD vocabulary lookup and LOD-based term disambiguation will be developed and evaluated, both quantitatively and with users. Secondly, EnviLOD will develop and evaluate intuitive user interface methods that can hide the complexities of the SPARQL semantic search language, while allowing environmental researchers to search successfully, using LOD vocabularies. A new British Library information discovery tool for environmental science, Envia, will be used as a case study to test the use of LOD vocabularies towards enhancing information discovery and management. Environmental consultants at HR Wallingford will collaborate as domain experts, providing feedback on how the semantic work undertaken here supports their work as environmental science practitioners and innovators.
This is a short overview of EnviLOD's progress by mid October 2012.
The key outcomes of this project are:
- User needs analysis, requirements gathering and use case definition.
- Open source tools for semantic enrichment with Linked Environment Data.
- A web-based interface for semantic search with Linked Environment Data.
- Quantitative and user-based evaluation results.
- A final report detailing the lessons learned.
- Publications, user workshop, and other dissemination and impact activities.
For further details see this blog post.
The British Library and HR Wallingford are working with us on this project. A complete list of staff involved is available on our blog.
Written Reports and Deliverables
- WP2: User Requirements
- WP2: User Feedback Report on the EnviLOD Semantic Search Interface K.Bontcheva, J. Kieniewicz, M. Wallis, N. Aswani, S. Andrews
- WP5: Evaluation Report on LOD-based Semantic Enrichment of Environmental Science Literature
Presentations from the EnviLOD User Workshop (25 January 2013)
- Information discovery in environmental sciences
- EnviLOD overview of the project: Linked Open Data, semantic enrichment, semantic search, our prototype search UI
Projects Collaborating With/Interested in EnviLOD Results
- The ViBRANT bio-diversity project and how the EnviLOD semantic search is relevant there
- The CHARMe project, which is considering a Linked Data approach, similar to EnviLOD. CHARMe aims to help users of climate data by providing more information in the form of "commentary metadata", which includes links to relevant publications, user feedback on datasets, discovered issues and so forth.
Project PI: Kalina Bontcheva
This work is supported by JISC