Open HaritzPuerto opened 3 years ago
Haritz – we absolutely need to use a real database going forward. That is, I believe, the conversation that Ismail wants to have with you.
From: Haritz Puerto notifications@github.com Sent: Tuesday, January 5, 2021 7:22 AM To: Coleridge-Initiative/RCGraph RCGraph@noreply.github.com Cc: Subscribed subscribed@noreply.github.com Subject: [Coleridge-Initiative/RCGraph] [Future Work/ Idea] Graph Database to store the data of the RCKG (#108)
If I remember correctly, I think @ernestogimeno https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_ernestogimeno&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=omwcNBUqPba9pikmkXZXk2bFQ7zxZPhI5OH9dd8lFDA&m=U0UeaCSQ2Cxl7eXrFTpdBoQxR6LYI-jMtUL4RYxQh50&s=sPoIpnq6hiGQXV994q6u7f06xsjHmK19lhPaVyAFIao&e= mentioned to me that the jsonld corpus format has some limitations that may affect the scalability so at some point we may need to use a real database.
A possible option is to use a graph database such as Neo4j. In this database, the information is stored in graph format so it might be suitable for our needs.
I just put this thought here for future discussion if we need to upgrade the jsonld format.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_Coleridge-2DInitiative_RCGraph_issues_108&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=omwcNBUqPba9pikmkXZXk2bFQ7zxZPhI5OH9dd8lFDA&m=U0UeaCSQ2Cxl7eXrFTpdBoQxR6LYI-jMtUL4RYxQh50&s=svTR1w-lKjBLJs6QwDsH2JndwBs1PiLpezxBbiwYWss&e=, or unsubscribe https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AD63AOZKV3BLNYISXPROTF3SYMAAFANCNFSM4VVAKJ7Q&d=DwMCaQ&c=slrrB7dE8n7gBJbeO0g-IQ&r=omwcNBUqPba9pikmkXZXk2bFQ7zxZPhI5OH9dd8lFDA&m=U0UeaCSQ2Cxl7eXrFTpdBoQxR6LYI-jMtUL4RYxQh50&s=KaJ57Il-Ck9H7ognQ38ik-KOKC1fbSfV_iVGlCG8i9w&e= .
If I remember correctly, I think @ernestogimeno mentioned to me that the jsonld corpus format has some limitations that may affect the scalability so at some point we may need to use a real database.
A possible option is to use a graph database such as Neo4j. In this database, the information is stored in graph format so it might be suitable for our needs.
I just put this thought here for future discussion if we need to upgrade the jsonld format.