I’ve been feeling somewhat bogged down in regards to project planning lately. I think a big issue is that my past programming projects have all been in class, with problems laid out, perfectly defined, by the text book or the prof. Making software in the real world isn’t that straightforward, and I think I am getting a taste of that lately.
This afternoon I met in person with Neal, my supervisor from LMSAL. He said that he too thought this project was somewhat ill-defined, and that it was not just my inexperience. This is quite a reflief to hear, actually.
In short, there seem to be two distinct but related objectives to this project: inference and faceted search. They both suggest fancy “semantic web” type ideas, but they also have more concrete and down-to-Earth examples. They both involve databases, which is something I should learn more about. The inference engine, using some sort of semantic magic on the database, could make inferences from the data, “if x and y, then z,” or more conretely, something like, “if active region A and active region B have the same ID number, then they are the same active region”. The faceted search would work something like on Amazon, except more along the lines of “scientists who were interested in these sorts of flares were also interested in these sorts of sun spots”. Also, faceted search could work something like the Zappos category search.
