327f4d4 · opencog-matrix

Changed files (2) +27 -2

opencog/Design-object-atomese.md +9 -2

…	…	@@ -4,7 +4,14 @@ Design thoughts about an object-oriented infrastructure, in Atomese.
4	4	nothing here yet, except for a sketch of ideas.
5	5
6	6	The below sketchs ideas (circa 2024) for an object-oriented variant of
7		Atomese.
	7	Atomese. It seems like it would not be hard to implement; it extends
	8	conventional Atomese in a minor way. Yet, it also feels insufficiently
	9	abstract. Over time, its become clear that the `ExecutionOutputLink` and
	10	`EvaluationLink` are problematic. They work, in the sense they were
	11	originally imagined to work, but they break up data flow in awkward
	12	ways. They provide the wrong kind of abstraction for a dataflow
	13	architecture. Thus, see [Design-similarity](Design-similarity.md)
	14	instead.
8	15
9	16	Motivation
10	17	----------
…	…	@@ -47,7 +54,7 @@ The `DefinedMethodLink` isn't really needed; one could just use a
47	54	`ListLink` and this would work just fine, today. But it's convenient to
48	55	have a special link type devoted to this task.
49	56
50		Invocation of the object-instance+method paure is "just like always", with
	57	Invocation of the object-instance+method pair is "just like always", with
51	58	the [ExecutionOutputLink](https://wiki.opencog.org/w/ExecutionOutputLink):
52	59	```
53	60	(cog-execute!

opencog/Design-similarity.md +18 -0

…	…	@@ -0,0 +1,18 @@
	1	Sesign notes for similarity measures
	2	====================================
	3	Design notes and ideas for similarity measures. Attempts to get at the
	4	core problem that the matrix API was invented to solve. Tries to do this
	5	at a deeper, more abstract and simpler way.
	6
	7	Motivation
	8	---------
	9	The current matrix API provides an extensive set of objects and
	10	functions for working with sparse vector embeddings. The vast selection
	11	is characteristic of what human programmers wnt and expect when
	12	attempting to write code to solve some problem. However, that same
	13	richness and variety impedes automatition and automatic construction.
	14
	15	Almost all of the matrix API was developed to support one generic
	16	problem: obtaining the similarity between two things. Thus, the task at
	17	hand is find some simpler, easier, more basic way of thinking about and
	18	working with similarity.