On this second week of coding I have made some progress on the next tasks: After this second week of coding, I have made some progress on learning more about my work on the GSoC.
Firstly, a meeting with my two mentors has helped me to decide what are my next steps on the project. They have recommended me to write an extension to this article describing all the architecture I plan to implement. So for the next three weeks, I will be centered on explaining the architecture with a graphic, and test if adding more examples to the dataset result on an improvement of the accuracy of the model.
On the second place, one of my mentors, Nandana, has helped me in the process of generating the dataset ready to train with weka. To do this process, we need to follow the next steps:
- Download the .tql files that are generated on the extraction process
- Reify those files into .ttl files that can be queried across a SPARQL endpoint
- The dataset is generated by doing queries against that triple store with a source code he provided me.
See you in the next week