Semantic Web and IPFS

[peacekeeper]

/re Trust-Exchange-An-Architecture-for-a-Permanent-Open-Trust-Network.md /re DecentralizedCooperationNeedsDecentralizedReputation.md /re Distributed-Trust-Systems-and-the-Kenyesian-Beauty-Contest.md.md /cc @jbenet /cc @harlantwood

Several papers here mention IPFS, and a lot of us are using semantic web technologies for representing profiles, relationships, tokens, and other data.

I'd be interested in mapping out how graph models such as RDF or XDI can be stored on top of IPFS' merkle DAG. I know some early thoughts on this topic exist, e.g.:

• @jbenet talks about "Interplanetary linked data": https://www.youtube.com/watch?v=OAkJAPS5yoQ minute 33
• Github discussion about IPFS-LD: https://github.com/ipfs/ipfs/issues/36
• Some implementation work: https://github.com/ipfs/go-ipld

[harlantwood]

Thanks @peacekeeper, great exploration!

Also, IPFS is hoping to add IPLD to their upcoming 0.4 release: https://github.com/ipfs/go-ipld/issues/14#issuecomment-151686128

[harlantwood]

/cc @dukedorje @aquabu @fractastical

[peacekeeper]

See @jbenet's contributions ipfs-keychain.md and ipfs-links.md

[fractastical]

+1 for this

[harlantwood]

Index/search capabilities of CRDTs being explored by IPFS. Some links:

• ipfs/notes#40
• ipfs/apps#12

[harlantwood]

Also related to the https://github.com/ipfs/archives project -- eg Freebase: https://github.com/ipfs/archives/issues/10

[elf-pavlik]

I tend to use RDF Dataset Normalization if storing RDF sources in filesystem, also Linked Data Signatures rely on it. The normalized g-text will always have the same checksum for particular g-snap

_source: https://www.w3.org/2011/rdf-wg/wiki/Graph_Terminology#Visual_Diagram_