Linked Data based Health Information Representation, Visualization and Retrieval System on the Semantic Web

• Author(s): Tilahun, Binyam Chakilu
• Date: 2013
• Persistent ID: http://hdl.handle.net/10362/9190
• Origin: Repositório Institucional da UNL
• Subject(s): Linked Open Data; Semantic web; Ontology; Health information; HIV; WHO
• Document Type: Master thesis
• Language: English
• Advisor(s): Kauppinen, Tomi; Keßler, Carsten; Painho, Marco Octávio Trindade
• Contributor(s): RUN

Description

Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies.

To better facilitate health information dissemination, using flexible ways to represent, query and visualize health data becomes increasingly important. Semantic Web technologies, which provide a common framework by allowing data to be shared and reused between applications, can be applied to the management of health data. Linked open data - a new semantic web standard to publish and link heterogonous data- allows not only human, but also machine to brows data in unlimited way. Through a use case of world health organization HIV data of sub Saharan Africa - which is severely affected by HIV epidemic, this thesis built a linked data based health information representation, querying and visualization system. All the data was represented with RDF, by interlinking it with other related datasets, which are already on the cloud. Over all, the system have more than 21,000 triples with a SPARQL endpoint; where users can download and use the data and – a SPARQL query interface where users can put different type of query and retrieve the result. Additionally, It has also a visualization interface where users can visualize the SPARQL result with a tool of their preference. For users who are not familiar with SPARQL queries, they can use the linked data search engine interface to search and browse the data. From this system we can depict that current linked open data technologies have a big potential to represent heterogonous health data in a flexible and reusable manner and they can serve in intelligent queries, which can support decision-making. However, in order to get the best from these technologies, improvements are needed both at the level of triple stores performance and domain-specific ontological vocabularies.