Wikipedia:Articles for deletion/Error-correcting codes with feedback

Error-correcting codes with feedback

Original research, not encyclopedic, reads like somebody's term paper Torc2 (talk) 23:18, 28 November 2007 (UTC)[reply]

week keep the article with a bit of work could be turned into a decent article or at least a stub. There is certainly an extensive literature on the subject from a variety of different authors.[1] --Salix alba (talk) 00:19, 30 November 2007 (UTC)[reply]
- Can you add any evidence of notability specifically for this topic? Torc2 (talk) 11:54, 30 November 2007 (UTC)[reply]
~~Delete.~~ No opinion. Yes it's original reasearch, although possibly peer reviewed, and in any case related to the author's PhD thesis. It is a chapter of a Springer book: Coding with Feedback and Searching with Lies. User Cdeppe, probably the author Christian Deppe, copied original LaTeX source code for a few paragraphs from that chapter into Wikipedia. Very likely he had heard that LaTeX markup works in Wikipedia and gave up when it turned out to be slightly more complicated. If there had been an obvious way for him to delete the article I suppose he would have done it himself. Since it's from a survey paper it might in principle make sense to spend the effort needed to wikify it. But under the circumstances I wouldn't trust that the author actually had the right to publish this excerpt under GFDL. I am not sure about the usual procedures. It should not be deleted in a way that makes it harder to create the page again. --Hans Adler (talk) 00:42, 30 November 2007 (UTC) (I changed my recommendation. I thought that the best use of editing resources would be to remove this and start the article from scratch when another expert becomes interested. While most arguments for keeping don't really convince me, on second thought it makes some sense to attract experts by keeping this article and linking it from error correcting codes. --Hans Adler (talk) 11:35, 30 November 2007 (UTC))[reply]
- It's NOT original research if it's been published in a peer-reviewed forum outside of Wikipedia first. "Original research", for the purposes of Wikipedia's policy forbidding it, is findings posted initially on Wikipedia that have not appeared first in refereed publications. Michael Hardy (talk) 02:20, 30 November 2007 (UTC)[reply]
  - Actually, it is. WP:COS and WP:NOT#OR specifically say this. Torc2 (talk) 02:31, 30 November 2007 (UTC)[reply]
    - No. Those say that even published authors can't write on their own knowledge without citations, but this is from a peer-reviewed book and so specifically not OR. I can't recall if the book is explicitly cited (it probably should be), but that's an editing issue not an OR issue. As to the article, I have no opinion at this time. CRGreathouse (t | c) 02:48, 30 November 2007 (UTC)[reply]
      - "If you have done primary research on a topic, publish your results in other venues such as peer-reviewed journals, other printed forms, or respected online sites, and Wikipedia will report about your work once it becomes part of accepted knowledge." - I think it's pretty clear that this is saying 'don't just copy and paste something you published, write about it,' and it's only permissible to cite the article. We can debate whether or not that meets the definition of the phrase "original research", but it's clear that it's the Wiki policy on original research is what covers this. Torc2 (talk) 04:04, 30 November 2007 (UTC)[reply]
        It says to publish it first before putting it on Wikipedia. Meaning wikipedia is not the primary venue for your research. There is absolutely nothing wrong with an author taking his own work and putting it up for GFDL. Everyone who clicks the "Save Page" button does precisely this. I hate to pull out this card, but you don't understand what original research means.
        
        You also seem to expect that a new article be perfect. Let me prove you absolutely wrong. Consider Richard III (1955 film). I started the article three years ago as a one-liner stub. It took 18 months but it achieved featured article status. By no means do you give up and delete an article just because it's not feature material. It's all the more reason to edit and improve it! Cburnett (talk) 06:20, 30 November 2007 (UTC)[reply]
        I understand what the WP:OR policy means, which is more relevant to this AfD. I also can comprehend unambiguous statements like the one I posted above; if you have an alternate interpretation of the rule I quoted that doesn't ignore the word "about" or interpret the word "cite" as "reprint", I'd be happy to hear it. And no, I don't expect a new article to be perfect, but I do expect one that's supposedly important and notable not to be totally orphaned and in the horrendous state this one was 75 days after it was created. As for Richard III (1955 film), ignoring the changes in policy and the amount of traffic Wikipedia has picked up since 2005, I would point out that your article effectively asserted its notability in a single line, was improved by others within a week, and was already linked to by another article before its creation. Comparing this topic to that article really don't really form a strong argument for keeping this. And I'm not giving up on this article because it's never going to be feature article material; I gave up because the article is unencyclopedic and doesn't assert or even seem to meet WP:N. It might be worth a line or two in another article, but it's not stand-alone material. Torc2 (talk) 07:56, 30 November 2007 (UTC)[reply]
        May I suggest that you stop this discussion, which really leads nowhere, and that we try to get a consense on the other open questions instead? I hereby excuse for having started this distraction by my unfortunate choice of words. Perhaps I shouldn't post in the middle of the night. --Hans Adler (talk) 10:20, 30 November 2007 (UTC)[reply]
        
        (Actually, another article linked to it before it existed because I linked it as I was disambiguating.) What you quoted says is this: "If A, publish A, and then A can be on wikipedia." It makes no differentiation between "publish" and "write about it", only that you can't publish to wikipedia as your primary venue. Period. The prose style used in a thesis or book is rarely in-tune with how wikipedia articles are written, but like I said right below is that {{cleanup}} is the way to handle such articles, not deletion. Cburnett 14:19, 30 November 2007 (UTC)[reply]
Keep. The poster of the chapter is reasonably believed to be the author (cdeppe == Christian Deppe) and so posting the content means the author put it up for GFDL. This is not the place to determine if the author still retained copyright or not. AFD is a bad answer for poorly/non-encyclopedic writing, that's what {{cleanup}} and regular editing is for. Cburnett (talk) 01:47, 30 November 2007 (UTC)[reply]
What in that article could you keep? The examples would have to go, and there's nothing notable about the concept in the article. The whole thing violates WP:NOT#GUIDE. I would have edited it if there was a single thing salvageable and notable for Wikipedia. Torc2 (talk) 02:22, 30 November 2007 (UTC)[reply]
- - Why would the examples have to go? You need to answer that if you want us to take that assertion seriously. Nothing notable in the concept? "Guide"?? It doesn't read like a textbook or how-to manual, and it's at least as notable as most articles on information theory or mathematics generally. Michael Hardy (talk) 02:43, 30 November 2007 (UTC)[reply]
    - It certainly doesn't assert notability. It has no sources or citations given and seemed to be, as has been established, just copied from elsewhere. "Alfréd Rényi reported the following story about the Jew Bar Kochba in 135 CE, who defended his fortress against the Romans" isn't any way to start off an encyclopedic example, and "Throughout this paper we shall call Carole and Paul the two players. This idea goes back to Spencer, who also explained: Paul corresponds to Paul Erd\"os, who always asked questions and Carole corresponds to an ORACLE, whose answers need to be wisely evaluated" isn't any way to end one. The first paragraph doesn't even tell the reader what the topic actually is. If it's so notable, why does "Error-correcting codes with feedback" and "Error-correcting codes with noiseless feedback" return a combined 25 Ghits? If it's so important, why is this article completely orphaned? Torc2 (talk) 03:15, 30 November 2007 (UTC)[reply]
      - Lots of articles on notable topics are completely orphaned when they're first created. Many of mine certainly were. An experienced Wikipedian immediately asks himself "Which articles should link to the one I just created?", then creates those links. The particular newbie Wikipedian we're looking at doesn't yet know that that should be done. And I get 389 google hits with the title of the article in quotation marks. Michael Hardy (talk) 03:35, 30 November 2007 (UTC)[reply]
        So you're saying that in the two-and-a-half months after this article, not one experienced editor looked and this and saw a need to integrate it into Wiki as a whole, yet the topic is critically notable? Are there any secondary sources that assert notability? Can you add them to the article and cite them? Torc2 (talk) 11:57, 30 November 2007 (UTC)[reply]
        You do get the inherit catch with your position? You can't get experienced editors to look at it if it's not linked; it's not linked because no experienced users have looked at it. Personally, I won't touch it until this AFD is done with to avoid having my time "deleted". Cburnett 14:19, 30 November 2007 (UTC)[reply]
Keep and clean up. If Renyi, Shannon, and Ulam thought it was notable, then it's notable. Michael Hardy (talk) 02:45, 30 November 2007 (UTC)[reply]
Keep The subject is notable, and the chances are that every one that posted here made use of technology that used this conecpt. The examples are valid in order to develop the concept, although I think the text needs a better introduction, since the reader is not necessarily committed to read the whole article, but may want a quick explaination up front. Graeme Bartlett (talk) 03:09, 30 November 2007 (UTC)[reply]
- If the concept is notable, why does searching for the article title on Google return almost no hits aside from the Wiki article itself? Is there anything about this that couldn't be summed up in one or two lines on the Error detection and correction or some other article? Torc2 (talk) 03:24, 30 November 2007 (UTC)[reply]
  - The article title in quotation marks (exact match) returns 389 google hits. Michael Hardy (talk) 03:36, 30 November 2007 (UTC)[reply]
    - I just double-checked the search and still got 15 results. That aside, 389 really isn't much better. Torc2 (talk) 03:57, 30 November 2007 (UTC)[reply]
      - Take out "num=100" like this. 389. Cburnett (talk) 06:07, 30 November 2007 (UTC)[reply]
        I just did the search (using your link and opening Google from scratch) from a totally different computer in a different city and got 15 links. I don't know why. If I click to show duplicates I still only get 38. In any case 15 vs. 389 for links is a pretty trivial difference.Torc2 (talk) 07:56, 30 November 2007 (UTC)[reply]
        I think that the subject of the article is sufficiently notable that it can be included. I am not sure that I fully understand the notability criteria in their application to scientific topics (is there no extra guideline on this?), but in any case if this article doesn't meet them then they have to be changed. Otherwise a scientific encyclopedia would inevitably be forked off Wikipedia. I get the 389 Google hits for "Error correcting codes with feedback". (And without drastic measures like reducing search to the US I can't reproduce anything like Torc2's numbers.) Considering that the phrase "regular polytope" (a former featured article) gets 11,400 that's not at all bad. We can't expect Associated Press to have an article on "Error correcting codes with feedback" once a year. Even if journalists did write about the topic they would very likely not use the exact phrase, judging it unacceptable for a general audience. --Hans Adler (talk) 11:26, 30 November 2007 (UTC)[reply]
        I just did the search using google.co.uk and got 393 hits initially; however, when I went to the next page, the sites found reduced back to 15 total. I also return 277,000 Ghits for "regular polytope". Torc2 (talk) 12:04, 30 November 2007 (UTC)[reply]
        I see now. Yes, it's the same for me, except I get only 11 (2 don't count because they refer to the Wikipedia article). But it's the same for "regular polytope." First you have to use the quotes to make it comparable: 11,300 hits. Then you have to click several times to get to the last results. Once you are there the number of hits is reduced to 370. It's still approximately the same ratio, 35, i.e. 1 1/2 orders of magnitude. Which is not at all much. My points were: 1. Low Google counts alone are not sufficient to come to a conclusion in an area where all Google counts are relatively low. 2. Long phrases and technical terminology are often covered somewhere without being mentioned literally. – But since I am not particularly interested in the outcome for this article I will now unwatch this page and concentrate on adding some content to Wikipedia, instead of participating further in a long discussion about whether or not because someone spent five minutes adding a copy of an original source we now have to preserve and expand it. --Hans Adler (talk) 12:48, 30 November 2007 (UTC)[reply]
        
        Regardless, the search litmus test doesn't assert verifiability as there are plenty of sources not indexed by the masterful google. Cburnett 14:19, 30 November 2007 (UTC)[reply]
- You still get pretty good results by searching "Error-correcting codes" feedback or even "Error-correcting codes * feedback". CRGreathouse (t | c) 13:19, 30 November 2007 (UTC)[reply]
Keep. No reasonable argument for deletion has been put forward. -- Dominus 14:06, 30 November 2007 (UTC)[reply]
- Comment. R.e.b.'s point below about the copyright issue is important. The article content should be replaced with something unobjectionable. -- Dominus 17:58, 30 November 2007 (UTC)[reply]
Keep. As the book chapter by Deppe is a survey article, covering work by Spencer, Ulam, etc., this suggests the subject is notable. There are also other surveys on this and related subjects, such as Searching with lies by Ray Hill, published in Surveys in Combinatorics 1995 (ISBN 0-521-49797-3). The article does need rewriting, categorising, to be linked to, and possibly also a better title, but I'm not sure what to suggest for one. The comments about original research are irrelevent: even if it is largely following Deppe's paper, as that's a survey it's reporting the work of others, so wouldn't be OR in any case. The google hits are also not particularly relevent: lack of google hits does not mean something is not notable, especially if there is more than one search term related to the topic. (If anyone wants to, they could google "Renyi-Ulam game" or "Ulam's searching game" instead.) Finally, the fact that it's lain untouched for a while is probably because very little links to it, so no-one had found it to do anything with it. --RFBailey 16:55, 30 November 2007 (UTC)[reply]
Delete. It's a well written introduction to an important and interesting topic, and is not OR. Unfortunately the copyright problem is more serious than people realize. It's a copy of a chapter in a book presumably contributed by the chapter's author. However according to the front of the book, the copyright is held not by the author, but by the Janos Bolyai mathematical society and Springer Verlag, so the author seems to have no right to post his own writing. (This is not a joke: in at least one legal case an author was sued by his publisher for posting his own book on the web.) I doubt the math society would object to its posting, but Springer might well do as they are trying to sell permission to view their own posted copy of the article. R.e.b. 17:36, 30 November 2007 (UTC)[reply]
- Comment I wonder if this problem can be solved by a suitable paraphrase. On many occasions people have posted material they've learned from a book and included a reference to the book. If it's deleted on copyright grounds then someone would have to replace it with another article of the same title covering more-or-less the same material. Michael Hardy 18:12, 30 November 2007 (UTC)[reply]
Keep. The arguement above about the applicability of "Original Research" distresses me; we're a Tower of Babel about editting philosophy. An author posting his own (published, peer-reviewed) work is no more evil than an artist contributing his own photographs to the public ___domain. On R.e.b.'s good point, the author quoting a small amount of his own marterial, citing the (copyright holding) source, can fall under Fair Use. The article can be improved with that in mind and does not need to be deleted IMO. Pete St.John 17:41, 30 November 2007 (UTC)[reply]
Keep, but as a redirect for now. I agree with R.e.b.'s comment, but not with the conclusion. This should never have come to AfD. The current content of the article is almost certainly a copyright violation, and should be removed: the article is not a critical commentary of the book in question, so is unlikely to be fair use. However, this does not mean that the subject of the article is not notable. The simplest thing to do is replace the current content by a redirect to Error-correcting codes until someone has the time and energy to write a new version which is not a copyvio, and also demonstrates notability in a clear manner. This approach preserves useful information in the edit history for anyone to take this up. Deleting the article destroys this information. I have some sympathy with the nominator, as I have in the past made the same mistake myself, by prodding articles which should simply have been replaced by redirects. I encourage others to take on board this option, as it saves a lot of time discussing issues like this at AfD. Geometry guy 18:28, 30 November 2007 (UTC)[reply]
Delete per R.e.b., as copyright violation. Geometry guy: while a redirect would not be harmful, we HAVE to delete copyvio content because otherwise it could be used to make still-violating derivative works. The author does not have the sole ownership of copyright in a published work; I know Springer-Verlag's copyright agreement it uses with authors, and the author agrees in it to assign the copyright to Springer, and cannot publish the work otherwise. So yeah, this is a copyright violation. Mango juice^talk 19:16, 30 November 2007 (UTC)[reply]