Thoughts on the current landscape of knowledge graph software:
Here's some of my favorite current knowledge graph software, it's geared toward advanced note-taking fanatics like me. 2300 + notes
Last updated: 20211207

Last updated: 20220218

From left to right

Traditional note-taking app + Knowledge graph:

Roam

✅ most polished of the three

❌paid & super expensive :/

❌ not open source,

❌ no mobile

for more on roam see this post

Roam: Turning Note-taking into a personal internet of you

Roam : Note taking app that turns your notes into a personal internet, generating a graph!Roam Research – A note taking tool for networked thought.As easy to use as a word document or bulleted list, and as powerful for finding, collecting, and connecting related ideas as a graph database. Collaborat…

Oran's BlogOran C

LogSec -  the new kid on the block

✅polished, but incomplete

❌small userbase :/ so no good plugins (

❌no mobile

link to website (LogSec )

Obsidian.MD

✅open source

✅free (free if you don't need a mobile)

✅ great user community

✅mobile, (UI is garbage, the best thing going for it is it exists)

❌garbage search this is a deal-breaker for me...

❌ sub-par UI, it's really close to being really good but it's not easy to use.

Obsidian's website here

My takeaways from using this software

Pretty knowledge graphs are great but they are not functional unless they have 3 things:

- render the same way, every time. ( deterministic rending) ( if the map wiggles or animates dynamically every time, like in obsidian it the graph becomes unparseable by my brain and thus unsearchable using my eyes, which defeats the purpose)

- Way to search the graph, by intuitively limiting visual complexity. when searching super large graphs it's essential

- Automatic node coloring, by tag or other programmatic methods

Weird but interesting projects:

Infranodus

is a totally different beast, it's not a note-taking software it's more of an exploratory research tool, which is quite hard to wrap my head around but it's got some really powerful search features that I think are best in class.
https://noduslabs.com/infranodus/

Bonus

Idea: Create a page rank algorithm
20211207

Create a page rank algorithm --- the one Google uses to map the internet  --- but apply it to knowledge graphs, I've not seen this done, and it would 100x improve the search capabilities.

Problem:
Currently, None of the above software uses page rank. and they all have a kinda garbage search ( LogSec is useable, but not exceptional)

core idea: knowledge maps + page rank = actually functional

# Final takeaway:

Knowledge graphs solve a fundamental scaling problem with large unstructured data.
But some really hard UI challenges haven't been solved for them to be a truly useful tool.
I have a gut feeling there is something groundbreaking here, it's just something I can't see yet.

I think the biggest challenge is a scale concerning data complexity, most of the projects I've seen work great for small datasets of 10-100 nodes, but fail when the visual complexity gets above the 300 node mark

Author

let oran_collins = ()=>{
return {
"github": "github.com/wisehackermonkey",
"email": "ɯoɔ˙ʃıɐɯƃ@ssǝuısnquɐɹo",
"Date Written":"d20211207"
}
console.log(oran_collins)