Frame (artificial intelligence) - by Seicnetsisnocni - rtnF

rtnF

SubscribeSign in

Frame (artificial intelligence)<br>Jun 27, 2026

Share

My first exposure to “frames” (the data structure) was when I took my “Knowledge Representation and Reasoning (KRR)” class in college. That class was an elective. I simply picked it because I was interested in it, and I wanted to pursue research in the field of knowledge management.<br>Here’s its definition from Wikipedia.

When I first encountered it, I didn’t really understand it. But there was an illustration in the KRR book that visualized what a “frame” is. It looked something like this.

An “entity” is described by several slots. Each slot has its own value. Each value has its own “type” (or data type), which could be an instance value or a link to another frame.<br>Years after passing that class, I stumbled upon something really similar to this concept of “frames”: Wikidata.

wdlite - Q686797<br>The terminology has changed a bit. Instead of an “entity,” it’s a “Wikidata item.” Instead of slots, there are “statements and properties.”<br>From there, I immediately became an avid Wikidata contributor.<br>It’s such a unique way to represent knowledge. Instead of representing it through natural text like in Wikipedia, in Wikidata we’re forced to abide by a certain schema and ontology. In some ways, it makes knowledge representation much more efficient (sometimes it’s easier to construct a knowledge graph triplet than to write it as a sentence, especially if your writing skills are poor). Yet at the same time, its current ontology is “not expressive enough” to represent complex knowledge.<br>In order to make the ontology “more expressive,” one can expand it by proposing new Wikidata properties. Doing so requires extensive discussion, especially to ensure that the proposed ontology remains in sync with, and does not conflict with, the existing ontology that has been built up over the years since Wikidata first went online. Which is... tedious.

Wikidata Property Proposal<br>Meanwhile, I see this kind of knowledge representation as quite an efficient way to represent knowledge. Maybe even for casual, personal use.<br>Well, Wikibase (the software behind Wikidata) is open source. One can simply install it on their personal machine and experience the full breadth of Wikidata in the privacy of a local machine. From there, you can use it for casual, personal knowledge management.<br>But I never tried installing Wikibase on my machine, despite wanting to use it to manage my own personal knowledge graph. I don’t have Docker installed on my PC.

Wikibase Suite Deploy documentation on GitHub.<br>So, I made it myself, building it from scratch using a tech stack that I already know: PHP.

index.php

Share

Discussion about this post<br>CommentsRestacks

TopLatestDiscussions

No posts

Ready for more?

Subscribe

© 2026 rtnF · Privacy ∙ Terms ∙ Collection notice<br>Start your SubstackGet the app<br>Substack is the home for great culture

This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts