Hello everyone,
It’s Linh Hoang, the intern from project 3. This week, my co-intern and I continue to focus on revising our YesWorkflow Data Model with our own vocabulary. Besides that, we also created an UML diagram to represent components of the model and how they connects to each other. The diagram is helpful since it is easier for outsiders to understand the model concept: what it includes, what are the relationship between the entities?, etc. We also had a long meeting with our advisors to discuss and ask for comments/feedbacks in order to improve the model. However, we also noticed that the more completed YW model is, the harder to map it with the existing ProONE model. This will be our challenge in the next few weeks, figuring out how to translate our own customized YW model to ProvONE model.
Besides working on designing the model, we also continue to test more SPARQL queries. Up till now, we have total 50 queries successfully run on Virtuoso and ARQ. These queries are both prospective and retrospective provenance. Some of them are across different RDF files (one file includes prospective data, one includes retrospective data). So far, our observations are:
- Comparing the two SPARQL querying tools: Virtuoso seems to have better performance than ARQ. It also supports more advanced SPARQL queries (such as advanced property path) than ARQ.
- Even though SPARQL is powerful and we are able to run all queries we had up till now. However, SPARQL queries can be getting complicated fairly quickly when user joins many “tables” (in RDF language, we call “objects”) together. Our previous querying methods such as XSB, DLV seems to be more simple (in terms of syntax).
I would love to discuss more details about SPARQL and its query capabilities with anyone who interested in. So please feel free to leave any comments you might have.
Thank you.
Linh