[03:21:04] Hi, I have a question [03:21:40] I understand that for a given wikidata item "Q42" say I can get all of its information in the form of a json at https://www.wikidata.org/wiki/Special:EntityData/Q42.json [03:22:14] My intention is to generate triples one for each claim made in the json of entity Q42 [03:22:54] The key values inside the response are pretty confusing mainsnak exists in some cases, doesn't in some other etc [06:18:42] AdityaAS: yes, Wikidata's JSON output is quite complex (because the data model is complex itself) [06:19:50] if you only want to do this for a bunch of entities, you can do this via SPARQL: "DESCRIBE wd:Q42" will give you all the triples it is involved in (see http://tinyurl.com/ya4n6wqz) [06:21:59] if you want to do this at large scale from the dumps, you can do it with Wikidata-Toolkit (but I think the RDF format it uses has diverged from the official format) https://github.com/Wikidata/Wikidata-Toolkit [07:25:10] pintoch: I have the rdf on my local server. I downloaded the Wikidata-Toolkit as well as Wikidata-Toolkit-Example repos but couldn't find an example of extracting all triples corresponding to an entity. Any advice? [07:37:44] you could adapt https://github.com/Wikidata/Wikidata-Toolkit/blob/master/wdtk-examples/src/main/java/org/wikidata/wdtk/examples/RdfSerializationExample.java [07:42:52] but if you already have the RDF dump, then you can just extract triples from it, right? [07:43:21] just index it with your favourite triple store and hit that [07:46:00] pintoch: I want triples corresponding to a particular entity [07:46:04] Say Q42 [07:46:15] yes I get that :) [07:46:47] Basically, we're building a knowledge graph for a specific domain from a corpus (Say a knowledge graph of all cisco devices) [07:47:10] This is our custom knowledge graph indexed on DGraph. We want to make it richer by including some triples from Wikidata [07:47:48] I want the triples in the format