SEWEBAR Proposal
From IKS Project
Contents |
Early Adoption and Validation Proposal of SEWEBAR
SEWEBAR (Semantic WEB Analytical Report) is a research project the goal of which is to study possibilities both of presenting results of data mining in the form of analytical reports and of disseminating resulting analytical reports through semantic web technologies. Ambition of this project is to become a framework covering the whole lifecycle of analytical reports: starting from the elicitation of knowledge from the domain expert, initialization of the data mining tasks, design of the analytical reports, to search over multiple mining results.
Use-Cases
Data Definition
1. CMS Admin/Manager defines data sources. Presently data sources are SPARQL endpoint and TMRAP webservices. In the extension it could be any service provided by FISE engine.
2. CMS Admin/Manager defines query templates and/or often repeated queries. These are strongly associated with data sources to be used with.
Knowledge Usage
1. Reporter builds query based on templates and the system sends request to designated data source.
2. Reporter uses query results inside CMS text editor (presently WYSIWYG editor, possibly Aloha Semantic editor)
Asynchronous Data Definition/Querying
Tasks in data mining can be very time consuming. In these cases the user would build the query and request the data source. The response should be received asynchronously so user is not active waiting.
Validation and Visibility
1. Internal validation by students of dataminig class
2. Deploy stable solution to our research partner, where the system is used by cardiologists.
3. Publishing CMS extensions to Joomla! community.
Performance
The only open performance issue is related to Asynchronous Data Definition and Querying. As far as this is not solved in FISE it is common objective we want to achieve.
