Talk:Semantic network

Semantic network and co-occurrence

In the social sciences people use the term semantic network to refer to co-occurrence networks (co-occurrence of words in units of text such as sentences). Would you mind if I add that under "Other examples" and link to co-occurrence networks? I would also add one sentence describing what co-occurrence networks are. --BenjaRoh (talk) 14:19, 27 May 2016 (UTC) — Preceding unsigned comment added by BenjaRoh (talk • contribs) 14:13, 27 May 2016 (UTC)

Adding 'Map Analysis' to the history section?

I am looking into content analysis techniques related to semantic network analysis, and I have found the related tradition of 'Map Analysis' which aims to produce semantic networks through both automated and hand coded analysis of relations in text.

I am considering adding a note about this research tradition to the history section. Does this belong here or elsewhere? Advice and opinions would be appreciated before I add a paragraph to the history section

Potential sources: Kathleen Carley, 1993. "Coding Choices for Textual Analysis: A Comparison of Content Analysis and Map Analysis" Sociological Methodology, Vol. 23 (1993), pp. 75-126 Kathleen Carley, 1993. "Extracting culture through textual analysis". Poetics 22 (1994) 291-312 --JonathanSchoots (talk) 01:57, 20 May 2016 (UTC)

Reword 1st paragraph

The first paragraph ("In 1909, Charles S. Peirce proposed a graphical notation...") is too similar to, as in, almost directly lifted from, the source it cites (AI, a Modern Approach, 3rd Ed., p. 454).

Agreed. It's not just "too similar", it's direct copyright infringement. I've removed the paragraph and tagged the article as such. I discovered the same fact independently, and it was immediately obvious. --50.98.255.28 (talk) 04:14, 6 November 2012 (UTC)

The paragraph is still included and I checked the original source and it is using the exact wording. I am removing the paragraph because of copyright infringement (and because it is not a particularly relevant piece of information) Akozlowski (talk) 22:16, 10 June 2016 (UTC)

Semantic Web

This article needs a short discussion of (and links to) the Semantic Web [[1]] and RDF [[2]] entries.

Also, DAML+OIL, Cyc and other existing ontologies.

semantic net = directed graph? really?

An Antonymy relationship is not directed. But if that is an important relationship of semantic nets, it anyways belongs to such a net. Thus, the semantic net cannot be directed. Correct?

Hence, I apply for the deletion of "directed" of "directed graph". --80.135.165.18 13:38, 25 February 2006 (UTC) (gneer, currently not logged in, since I don't have my pwd at hand)

Many of the relationships are directed (isA for example). Non-directed relationships can be thought of as bidirectional. Perhaps we should mention that somewhere? Drevicko (talk) 00:18, 7 October 2008 (UTC)

Incorrect use of "Vertebrate" in image

"Vertebrate" in the image used as an illustration of a symantic network, base on the stated relation "has an", should be "Vertebra" because a vertebrate IS an animal with a spine--an animal with a spine is not something a mammal HAS--and a vertebra, which a mammal does have, is a bone of the spine. Alternatively, although it would reduce the number of types of relationships illustrated, the relation between Mammal and Vertebrate could be rewritten as "is a" rather than "has a" to indicate that a mammal is a vertebrate. In this latter case, the concept Animal could further be pointed to by the concept Vertebrate since the concept Vertebrate also has the semantic relation "is an" with respect to the concept Animal.--71.142.61.168 17:50, 17 June 2007 (UTC)

Thanks for pointing that out. I will try to correct the diagram changing Vertebrate->Vertebra (I think your second proposal would just look too cluttered - the diagram should just provide an illustration of semantic nets rather than a detailed structure on animals).--Konstable 12:08, 18 June 2007 (UTC)

Unclear

I don't understand this sentence: "Such networks involve fairly loose semantic associations that are nonetheless useful for human browsing." I think it is too judgmental about WordNet and it doesn't add information. Why "fairly loose"? —Preceding unsigned comment added by Sandman2007 (talk • contribs) 18:03, 27 November 2007 (UTC)

Semantic Web series

I added the macro for the semantic web series - it contains important links to related concepts such as ontologies.

However the page formatting has suffererd: the 'banner' on the right hand side with the semantic web links vies for the same place on the page as the image relating to the introductory text. I have left the image in 'pole position' - the banner is pushed down the page. This does not seem ideal to me though, but I'm not sure how to fix it, nor am I sure if putting the side by side would be better? Drevicko (talk) 00:18, 7 October 2008 (UTC)

Possible Reorganisation ?

History

Examples

WordNet

Gellish

Lexipedia

Semantic Network Construction

(That piece between WordNet and Gellish would go here)

etc.

Mnbundle (talk) 13:14, 14 July 2009 (UTC)

Role of Quillian?

It seems that the history substantially understates the role of Quillian in the development of semantic nets. Indeed, I often hear them referred to as "Quillian's semantic nets" (or Quillian's semantic network theory) and his thesis is where I thought the term originated. Sowa's article treats "semantic nets" more generally, but event there, Quillian is called out by name.

Actually, the whole history section seems very tendentious at at odds with the standard views (at least in computer science). (The standard views, as I understand them, are 1) Quillian is the originator of semantic networks per se and 2) the Sowaian view that all sorts of graphical representations of knowledge are semantic nets so they have a really long history, but that Quillian is in the mix. (I take the latter to be a minority view, actually.) —Preceding unsigned comment added by 130.88.198.12 (talk) 12:29, 17 November 2009 (UTC)

You seem to know quite a lot about semantic networks... It would be great to use your knowledge to improve this and other related articles, since as you say it is still very very far from having high quality content and a neutral point of view. If you finally decide to collaborate with wikipedia and you have any questions about how does wikipedia work feel free to ask them at my talk page or even here. Best regards. I hope to see you soon.--Garrondo (talk) 12:36, 17 November 2009 (UTC)