2 ms·
I am just curious, but what do you feel defining these three common terms, surely all well known to anyone who'd read HN, adds to the discussion?
by lproven 3y ago
I am just curious, but what do you feel defining these three common terms, surely all well known to anyone who'd read HN, adds to the discussion?
- westurner 3y agoLinks are edges. When there are URL links, there are better edges that cost less to find than NLP tasks like conceptual entity recognition. Because I've added edges from hn_uri to wikipedia_uris; I, others, and search engines could discover the hn_uris relevant to the concept URIs (by graph pattern search) W3C SPARQL is a graph pattern query language. Some triple graph patterns enabled by there being links in the comments here in HN: ?s ?p ?o. ?s linksTo ?o. ?s linksTo <wikipedia_uri> . Also, recently, for example, because I linked to "methane" on Wikipedia unnecessarily, and then checked the "See also" section, and then saw the "Global Methane Initiative" page, I realized that there was an opportunity to forward the latest methane emissions tech to GMI (which I learned about by reading comparatively mundane background information from Wikipedia, which has Concept URIs which are reused by DBpedia which contributes to the Linked Open Data cloud for indeed what returns)
- westurner 3y ago(I haven't even yet written a "parse into RDF and load into triplestore/quadstore" probably because just full text search over teh hnlog with Ctrl-F is usually good enough to find the citation I'm trying to recall; but search engines do index URL edges with some magnitude)