Hacker News Reader

Comment by productceo | Hacker News Reader

I grew up hearing a lot of buzz for concept maps (or mind maps as I heard). I never found them to be as powerful as advertised.

A concept map is slightly easier than writing in bulletpoints because it can ignore structure. But this makes it incredibly difficult for anyone, including myself, to read after it is written.

At best, I use a concept map as an intermediary stage for figuring out how to write in bulletpoints, write in paragraphs, or draw in diagrams. This intermediary stage never lasts overnight for me, and I discard the concept map immediately.

View on HN

Do you mean real mind maps and real concept maps, or one of the many bastardizations?

Mind maps are very different from concept maps, except that they both arrange boxes and lines on paper.

If you look at Buzan's Mind mapping book, you'll find that he has a definition that virtually no mind map in common use adheres to.

(https://www.2uo.de/Books/the-mind-map-book/ is very incomplete, just a few notes, but look especially at "central image", "branches varying in thickness", "Only one word per branch" and "lines only as long as words")

Similarly, at least the first twenty videos on YouTube I've seen promising to show concept mapping aren't concept maps at all. They are "spider diagrams" (as Buzan calls them), usually with artsy ambitions. And for some reason, many of those were from nurses-in-learning – the term "concept mapping" seems to be very much en vogue in that field.

I'm a very textual person myself, working with endlessly tab-indented text files, but I'm currently trying to look over the fence to the more graphical approaches.

I didn't dig deep enough to know the difference between the "real" ones and the illegitimate ones.

My current approach is, don't fix what is not broken. If I'm happy with the system I have now, let it stay until there is a need to change it.

I do wish you luck on your exploration of graphical approaches though. Let me know if you find them useful. :)

I’m a textual person 90% of the time for the speed and flexibility. But I’ve found visualizations of my textual thoughts — after writing them — in the form of a mind map to be quite valuable in some circumstances. Mostly to communicate vision in a way that’s not top down.

At the end of the data, an endlessly tab indented text file is no different than a mind map. Instead of text files I use infinite list tools such as Dynalist, which have a great feature to convert your list instantly into alternative views such as mind maps.

https://help.dynalist.io/article/84-switch-to-another-view

There is a fundamental difference between Mind Maps and Concept Maps (although both are pen-and-paper based visualisation techniques from the 70s, that have proven helpful for learning):

Mind Maps have a hierarchical structure like trees. One topic in the middle, branches and subbranches from there. They are good for: brainstorming, outlining topics, clustering. They suck for: showing interrelations between concepts.

Concept Maps have a network structure. They are (labelled, directed) node-and-link diagrams. They are good for showing interrelations, influences, more complex structures. They suck for: quick and easy outlining, brainstorming, clustering.

What both have in common: They don't scale. Maps with 100 items or more become hard to read and deal with.

The only approach I know of that unites hierarchy and interrelations while maintaining the core benefits of stable spatial layouts is the "iMapping" approach taken by Infinity Maps. Plus it scales. You can easily build maps with thousands of items that are still clearly readable.

https://infinitymaps.io/en/

Here is an impression of Infinity Maps: https://www.youtube.com/watch?v=5Wfq--1t8l0

Thanks for that -- it looks really interesting. (It's crazy that Infinity Maps doesn't seem to have a single screenshot of the actual interface on their website, or at least I couldn't fine one after clicking around for a while; I can't imagine how they attract new customers.)

Eastgate Tinderbox allows a similar nested spatial model.

That’s right. Some differences to tinderbox though: Infinity Maps is - Web instead of Mac only - collaborative (real-time) - lets you zoom smoothly through the hierarchy and see several levels at once

An older app that is similar is The Brain. It allows you to move in and out of topics and does a decent job of dealing with large numbers of items. I used it a lot in the early 2000s and it's still around. https://www.thebrain.com/

("Mind maps are similar in structure to concept maps, developed by learning experts in the 1970s, but differ in that mind maps are simplified by focusing around a single central key concept."):

https://en.m.wikipedia.org/wiki/Mind_map

Perfect one sentence summary. I wasted a minute or two searching online for the difference between the two, but I should have read your comment first! Thanks for communicating precisely and concisely.

While scoping is necessary and useful, I hadn't ever learned that there were categorical restrictions upon mind map graphs, definitionally

Centrality, Components, Cliques might be helpful for general network analysis: https://networkx.org/documentation/stable/reference/algorith...

The difference between a URI and a URL; a URL is supposed to be dereferenceable and retrievable.

Named Graphs are composed of triples:

  {named_graph_uri, subject_uri, predicate_uri, value_uri, [datatype_uri, language_uri]}

  {g, s, p, o, [d, l]}

E.g. Rdflib supports various stores for and representations of RDF: https://en.wikipedia.org/wiki/RDFLib

Though, general systems are more complex than the average [concept] network: there are nonlinearities in so many of the observed relations between components that trees and DAGs are laughably insufficient; systems descriptions require more than naievely-acyclical graphs without nonlinearity in independent relations even.

Graph (disambiguation) https://en.wikipedia.org/wiki/Graph

Glossary of systems theory https://en.wikipedia.org/wiki/Glossary_of_systems_theory

How best to describe nonlinear relations without a MultiDiGraph that bundles edges together by [QFT] field, because they're objectively not statistically independent?

s/value_uri/object/, which may be a URI:

  {named_graph_uri, subject_uri, predicate_uri, object, [datatype_uri, language_uri]}

  {g, s, p, o, [d], [l]}

A mindmap, where e.g. g=example.org/mind-maps/2022#our-mind-map and "a" is "xsd:type":

  @base <_#> .
  @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
  @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
  @prefix schema: <https://schema.org/> .

  :Node, a, rdfs:Class .
  :linksTo, a, rdfs:Property .

  :Shapes, a, :Node .

  :Shapes, :linksTo, :Rectangle .
  :Rectangle, a, :Node .

  :Rectangle, :linksTo, :Square .
  :Square, a, :Node .

  :Square, rdfs:label, "Square" .
  :Square, schema:name, "Square" .

  :Rectangle, rdfs:subClassOf, :Shape .
  :Square, rdfs:subClassOf, :Rectangle .

https://en.wikipedia.org/wiki/Notation3#Comparison_of_Notati...

https://www.w3.org/TR/turtle/#sec-intro

Concept maps are not mind maps, concept maps work in a different way. I'll just offer a self-reference about concept maps [1] since there I have an explanation of how they work I like.

[1]: https://mostlymaths.net/2021/10/concept-maps.html/

I recently read about a discovery that may explain that preference.

Psychologists have realized that it is easier for humans to remember a list than a set. That is, memorizing objects AND the order in which they appear is easier than just memorizing the objects.

A lot of our memory is based on sequences. We may think of it as extra information that clutter our memory capacity, but it is actually not the case: have things in a given order, in a given place, in a given color, and it is easier to memorize than if you remove information.

IME an arbitrary set of objects is best recalled as a mind-palace arranged in 2D space. Think about the typical display of buttons or labeled icons in our GUI's. That kind of visualization is inherently memorable, even more so than listing the same content in some arbitrary sequence.

Perhaps someone can write some GPT-3 based tool to convert these maps into linear text.

This would be great for writing documentation.

> I grew up hearing a lot of buzz for concept maps (or mind maps as I heard). I never found them to be as powerful as advertised.

Agreed. I find they constrain thoughts far more than just a simple list. However, I know that they are helpful for some people.

Reply on news.ycombinator.com