Talk:Hierarchical classification
 | This article is rated Stub-class on Wikipedia's content assessment scale. |
No sources cited
editThis page doesn't seem to cite its sources... (Apologies if I'm reporting this in the wrong place, I'm new to the Wiki.) — Preceding unsigned comment added by 81.86.160.116 (talk) 00:30, 25 August 2007 (UTC+9)
- That's OK to discuss here, yes --mcld (talk) 10:06, 23 September 2009 (UTC)
- Just want to document, I'm an expert on this topic. I agree the article is currently a mess, no sources. I'm editing it now and plan to post a very different version with major edits and references in a few days. If anyone has any other issues on this topic please document them here so we can discuss. --MadScientistX11 (talk) 12:45, 12 July 2014 (UTC)
Not done Realized this article is not about what I originally thought (shows how bad it currently is that it took me this long to realize what it was trying to say). So I'm not going to edit it after all, it's not my area of expertise. See comment below for more info. --MadScientistX11 (talk) 17:37, 12 July 2014 (UTC)
- Just want to document, I'm an expert on this topic. I agree the article is currently a mess, no sources. I'm editing it now and plan to post a very different version with major edits and references in a few days. If anyone has any other issues on this topic please document them here so we can discuss. --MadScientistX11 (talk) 12:45, 12 July 2014 (UTC)
"External links to people" - ???
editWhy "External links to people"? Surely we should be listing techniques rather than people? --mcld (talk) 10:06, 23 September 2009 (UTC)
Similar Models section
editFrom my first read this section is OR. There are no references and I plan to just delete the whole section unless someone can provide a reference and defend it. I see what the person is saying but they are just drawing some very general analogies and similarities between various forms of computation and what's more I don't know anyone who has said those kinds of things in print and I'm well read on the topic, I'm doing work with Protege right now -- see: http://protege.stanford.edu/ -- which is the state of the art as of now (2014) on supporting these types of classifiers. --MadScientistX11 (talk) 12:51, 12 July 2014 (UTC)
- I got some good advice from another editor (thanks User:StarryGrandma) in the teahouse and I realized that this article isn't about what I thought it was. I assumed this was about deductive classifers such as implemented in languages like KL-ONE but I realize now there is a whole other topic called Hierarchical Classification that uses a connectionist, neural net approach rather than the formal logic approach that KL-ONE style classifiers use. So, this is an Emily Latella moment... never mind. I do think this article needs LOTS of work but I'm not the right person to do it and I'm going to create a new separate articles on deductive classifiers. My advice for this article though is ditch all the "it's kind of like this, and it's kind of like that" talk, that's just OR and vague, say what it is clearly, how it's used, and the basic ideas of how it works. I'm a technical guy and I can't get any of this from the current article. --MadScientistX11 (talk) 17:32, 12 July 2014 (UTC)
- I just re-read this section a bit more carefully. This sentence is just flat out wrong: "The generally accepted view of the brain today is that the brain is a generic pattern machine that works to abstract information again and again until it relates to a broad stored concept" First of all our understanding of the mind/brain is so immature there are very few substantial ideas that are "broadly accepted". But if anything, at least from the authors I've read recently such as Steven Pinker and the evolutionary psychology people the trend is just toward the opposite view, away from "the blank slate", away from a general pattern matcher that people like the early AI researchers imagined and toward distributed modules with diverse functions. But in any case this is supposed to be an artile about an algorithm to classify text not neuroscience or mind/brain psychology. And the rest of that paragraph is equally irrelevant and incoherent. --MadScientistX11 (talk) 23:48, 25 July 2014 (UTC)
- Just wanted to document that I have deleted the entire "Similar Models" section for reasons documented above. --MadScientistX11 (talk) 13:58, 31 July 2014 (UTC)
- I just re-read this section a bit more carefully. This sentence is just flat out wrong: "The generally accepted view of the brain today is that the brain is a generic pattern machine that works to abstract information again and again until it relates to a broad stored concept" First of all our understanding of the mind/brain is so immature there are very few substantial ideas that are "broadly accepted". But if anything, at least from the authors I've read recently such as Steven Pinker and the evolutionary psychology people the trend is just toward the opposite view, away from "the blank slate", away from a general pattern matcher that people like the early AI researchers imagined and toward distributed modules with diverse functions. But in any case this is supposed to be an artile about an algorithm to classify text not neuroscience or mind/brain psychology. And the rest of that paragraph is equally irrelevant and incoherent. --MadScientistX11 (talk) 23:48, 25 July 2014 (UTC)
What is this article about?
editI'm not sure that a specific use in AI is the real meaning of "Hierarchical classification".
Sources for other uses include:
- https://www.canr.msu.edu/resources/hierarchical-classification
- https://arxiv.org/abs/2005.06692
- https://towardsdatascience.com/https-medium-com-noa-weiss-the-hitchhikers-guide-to-hierarchical-classification-f8428ea1e076
PamD 07:45, 2 April 2022 (UTC)
- Agreed. See: WhatLinksHere/Hierarchical classification and the deletion log
- The article used to have much more information (example revision from 2010), but nearly everything was eventually removed due to a lack of references.
- (It seems to have been moved from Hierarchical classifier, which probably didn’t have as much overlap with other vernacular usages.)
- It does need to be expanded, and your links may be promising starting points.
- See also:
- - Jim Grisham (talk) 00:24, 23 July 2022 (UTC)
- The AI-related usage, included its linked Wikidata info, should probably be a separate article (e.g. Hierarchical classification (Artificial intelligence) / Hierarchical classification (Machine learning)), referenced from a Disambiguation page at Hierarchical classification (or at Hierarchical classification, if consensus is that the traditional use of the term is not the ‘primary topic’).
- - Jim Grisham (talk) 00:33, 23 July 2022 (UTC)