mic_none

Wikipedia:Articles for deletion/Natural Constraint Language Source: en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Natural_Constraint_Language

The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was delete. There have been a number of "keep"-!votes but they fail to make a policy-based argument that this topic is notable. SoWhy 17:46, 18 June 2011 (UTC)[reply]

Natural Constraint Language

[edit]
Natural Constraint Language (edit | talk | history | protect | delete | links | watch | logs | views) – (View log)
(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

No independent secondary sources to establish notability as required by WP:GNG. A search suggests independent sources may not exist. The only contributors to the article are WP:SPAs. Msnicki (talk) 13:27, 8 June 2011 (UTC)[reply]

Note: This debate has been included in the list of Computing-related deletion discussions.
It's certainly highly technical. But generally informative is another matter. I think I know enough about programming languages and compilers and what an NP-hard problem is that if an article on one these topics has been written WP:NOT#JARGON "for everyday readers, not for academics" that after reading it, I should be able guess what it's about. But I have no idea what this thing is or what it does. This article told me nothing. Do you write programs in this language and do they get compiled and run? What do they do? Who can tell. I'm guessing it's yet another academic (in this case a researcher at Microsoft) self-promoting his little-known work. WP:NOT PAPERS, WP:NOTPROMOTION, WP:INDISCRIMINATE Msnicki (talk) 15:22, 8 June 2011 (UTC)[reply]
I was able to take a bit more out of it. (I must be a very clever fellow!) I gathered that this theory or research had to do with solving complex problems involving constraints (how do I paint the map with a limited palette, so that no border has the same color on either side? What's the best way from here to the zoo, avoiding traffic jams, stop lights, and extra miles, with each factor weighted?) These problems are mathematically hard, from what I'm told. Solving them is one of the few things humans usually do better with than computers. I don't know enough about the field to suggest a merger subject, but I did want to point out that this seems to be well written technical material of the sort we ought to save somehow if possible. The two articles do seem to be mergeable, though. - Smerdis of Tlön - killing the human spirit since 2003! 20:29, 8 June 2011 (UTC)[reply]
Well, now you're just talking about what an NP or NP-hard problem is, not anything this article could have contributed to your knowledge. But what does this Natural Constraint Language stuff have to do with any of that? Do you describe one of these problems and have it tell you if a solution is computable? Does it search for good solutions if perfect ones are too expensive? Or is this just a notation and perhaps an algebra for working these kinds of things on a whiteboard? Who can tell. Msnicki (talk) 20:37, 8 June 2011 (UTC)[reply]
People can understand very well that if a mathematician describes an industrial problem naturally in conventional mathematical logic and a computer language can recognize it and solve it efficiently, it is fantastic. NCL can do this. NCL aims to solve industrial problems such as logistics optimization, production scheduling, human resources optimization and other problems. So this is a very hard research and very interesting, though it may be a long and lonely work. SophiePaul (talk) 12:10, 9 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
  • Delete. Keep. The article is based on primary sources by a single author, and therefore doesn't satisfy WP:GNG. I spent some time searching Web of Science for more sources, but found nothing useful; I found only two citations of Dr Zhou's articles by other authors, and those papers didn't include any significant discussion of NCL. It seems that a company called Enginest uses NCL, but I don't think that counts as an independent source since Dr Zhou seems to be affiliated with them. I agree with Smerdis of Tlön that the article looks interesting and it would be nice to retain this content. However, without seeing an independent source, ideally a secondary source, it's hard to justify keeping it. Jowa fan (talk) 06:53, 9 June 2011 (UTC) In the light of the discussion below, and additions to the article, I now believe that the necessary sources exist. The article could still use some work; an explanation of which reference supports which parts of the article would be helpful. Jowa fan (talk) 13:10, 14 June 2011 (UTC)[reply]
There seemed to be some problem with the internet (slow for us to respond). Please try searching on "www.google.com" by typing "NCL Constraint Language". If you only type "NCL" or "Natural Constraint Language", it may not give you sufficient information because there are quite some others also named "NCL". There are many reliable and independent sources citing NCL, including users of NCL, research papers citing or using NCL, and research center links, etc. Even in Mainland China a Book "自然约束语言"(in Chinese) was published in 2009. A book "The NCL Natural Constraint Language" in English will also be published soon. Please note that these Wikipedia articles need to be further developed and it takes time. NCL and "Mixed Set Programming" are purely scientific concepts and research on them are highly advanced. The research work on NCL has been very hard since more than 12 years. This technology has been proved to be very successful in industry. So please help us to improve the quality of these articles. At least please do not delete them. MERCI. SophiePaul (talk) 13:53, 9 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
Using google translate to view that link, I think the author of the book is the same Dr Zhou. Can you name a book or paper on this subject by a different author? Jowa fan (talk) 02:58, 10 June 2011 (UTC)[reply]
"Related works" is added to indicate relevant works, including POEM and a book on an experimental system based on POEM. By the way, Google Scholar search for "J. Zhou: Introduction to the constraint language NCL" can tell more relevant works and citations. Thanks for your attention. SophiePaul (talk) 14:22, 10 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
  • Please do not delete. 1) Please retain the articles "Natural Constraint Language" and "Mixed Set Programming". They are about new subjects, fundamental and are significant for industrial applications. Independent research work since more than 12 years by J.Zhou (who is not in academy, no need for academic promotion) should be respected scientifically. A search for "NCL Constraint Language" on Google can tell that there are many sources about NCL and MSP. 2) Please do not merge NCL and MSP. They are different subjects. "Natural Constraint Language" is about a computer programming language for modeling and solving problems. "Mixed Set Programming" is about an algorithmic framework. 3) For NCL, there is a journal paper "J.Zhou. Introduction to the constraint language NCL. JLP 45(1-3): 71-103(2000)". A book "The NCL Natural Constraint Language" in English will be published soon. These two Wikipedia articles will also be further developed to provide readers with detailed knowledge. Merci. SophiePaul (talk) 11:46, 9 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
In reply to point (1): I did search on Google before I made my comments above. What we need is sources that are reliable and independent. I didn't see anything that looked suitable. Have you read WP:GNG carefully? If you can provide links to appropriate sources according to those guidelines, then it may be possible to keep the article. Jowa fan (talk) 12:14, 9 June 2011 (UTC)[reply]
See Wikipedia:Articles for deletion/POEM (software) for an example of an independent source that's also relevant here. I think we still need at least one more independent source. Jowa fan (talk) 03:16, 10 June 2011 (UTC)[reply]
"Related works" is added to indicate relevant works, including POEM and a book on an experimental system based on POEM. It seems that recently some other researchers are working toward similar objectives, but they are still quite different from NCL. Please be understanding that further development on these Wikipedia articles is being prepared. It takes some time. Thanks. SophiePaul (talk) 14:22, 10 June 2011 (UTC)— SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
The book by Li appears to be about logistics in general. (Sorry, I don't have access to a copy of the book, I'm just going by Google's translation of the web page.) Does it contain significant coverage of NCL? It's not clear that the other related works have anything to do with NCL at all. More information is needed! Jowa fan (talk) 03:11, 12 June 2011 (UTC)[reply]
I am informed that the book is about logistics optimization modules developed using POEM. All math models (vehicle routing, production scheduling, etc) are programmed in NCL and run in POEM. So that work is closely-related to POEM/NCL. The 3rd and 4th (independent) related works cite Dr Zhou's work (result of 12 years ago) and they target similar objectives. A question might be: if a subject is extremely complex and the research work on it is terribly hard and nobody else works on it, Wikipedia will reject an article on such a result? NCL involves techniques in different scientific fields: programming language (formalism/grammar theory/compiler/parser/pattern recognition), operations research (combinatorial optimization/complexity theory/algorithm), logic programming (first-order logic/numerical reasoning/naive set theory),... SophiePaul (talk) 10:16, 14 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
With some others, I have been trying my best to improve according to your comments. And such improvements will continue. Please let me know your further advices if any. By the way, I think when I initiated the articles, I have paid enough attention not to include any "delicate" information (e.g., a company name) so that it is not considered a "publicity". In fact, if Wikipedia could host a document I would have uploaded Dr Zhou's papers into Wikipedia's space instead of linking to the current sites. SophiePaul (talk) 10:16, 14 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
Is "poorly written" really a reason for deletion? Jowa fan (talk) 13:10, 14 June 2011 (UTC)[reply]
If the article could be improved perhaps not. However, SophiePaul seems to be the only person with access to independent sources that are claimed to exists. Given his peacock term filled language at this AfD and incomprehensible writing skills demonstrated in the current articles, I have no doubt he will fail at this task. —Ruud 16:41, 15 June 2011 (UTC)[reply]
It should be kept The article is improved by adding more reliable and independent sources. 1) Dr Zhou had quitted academy and had worked in industry since 1999 before his NCL paper was published. 2) (Zhou 1996; Zhou 1997), describing a pre-prototype of NCL and scheduling, is only a tiny part of (Zhou 2000). 3) "Ridiculously", (Zhou 1996; Zhou 1997) receives much more attention (16+29 = 45 citations in GS). It is even cited by Prolog's inventor Alain Colmerauer's paper. See comments below. 4) Recent NCL (Zhou 2009) is far more advanced than (Zhou 2000). 5) A book in English has already be finished. And it certainly will help develop the article. 6) Hope this answers your comments. SophiePaul (talk) 17:29, 15 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
You're only supposed to WP:!VOTE once, SophiePaul. Msnicki (talk) 12:59, 14 June 2011 (UTC)[reply]
  • Question. What sources are now being offered in support of notability? Since being nominated for AfD, I see that all kinds of supposed sources have been added. But what is this junk? I'm pretty darn sure that Don Knuth, Noam Chomsky and Aho and Ullman did not have anything to say about NCL. This is nonsense. You really only need 2 decent independent articles or mentions to establish notability and I still don't see them. Instead, all I see is a lot dust being thrown in the air. What are the ACTUAL sources offered in support of notability? I think it's still zero. Msnicki (talk) 13:28, 14 June 2011 (UTC)[reply]
I too am mystified by the list of "references". Certainly the article as a whole needs some work. The actual sources I'm interested in are the ones under the heading "related works", specifically the book by Li and the articles Martín et al 2009 and Flener and Pearson 2004. See SophiePaul's comment above, indented and dated 10:16, 14 June 2011 (UTC). Jowa fan (talk) 01:42, 15 June 2011 (UTC)[reply]
    • Answer The only source mentioning NCL , MSP and POEM seems to be the papers by it's author/inventor J. Zhou. The other references are general computer science text not directly on the subject. —Ruud 16:29, 15 June 2011 (UTC)[reply]
    • Please read Improved according to your advices. Please see added related works; they are from Google Scholar by searching for "Jianyang Zhou". 1) (Zhou 1996; Zhou 1997) presents a pre-prototype of NCL for scheduling. That work is a tiny part of (Zhou 2000). 2) (Bleuzen-Guernalec and Alain Colmerauer 2000) is on "Sorting Constraint" (sorting is fundamental for computer sicence); (Beldiceanu 2000) is on "Global Constraints"; (van Emden 1999) and (Jaulin et al. 2001) are on "Interval Constraints"; they cite (Zhou 1996; Zhou 1997). 3) (Martín et al. 2009) and (Flener and Pearson 2004) are on "constraint systems"; they cite (Zhou 2000). 4) (Li 2009) is a book on an experimental system based on POEM/NCL. It presents optimization models programmed in NCL and it helps in teaching students about logistics and optimization. 5) I believe: what is important is the scientific value, not a publication. Dr Zhou did not seek to publish during the past 12 years though around his PhD his publication is remarkable. However, a book in English has just been finished. SophiePaul (talk) 17:29, 15 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
      • The paper by Bleuzen-Guernalec and Colmerauer (preview of first page) doesn't say more than "More recently Jianyang Zhou [5] has solved well known difficult job-shop scheduling problems by introducing a sortedness constraint with 3n variables, then extra variables being used for making explicit a permutation linking the xi’s." It does not mention NCL. Can you name any independent papers which do? —Ruud 17:40, 15 June 2011 (UTC)[reply]
      • While Zhou's work has been published in well-know computer science journals, including the Journal of Logic Programming, citations of his work are minimal. I don't think this justifies having three separate articles on his work. I could imaging it might be briefly mentioned in an overview article on constraint logic programming, but I don't think this is notable enough for an independent article. —Ruud 17:54, 15 June 2011 (UTC)[reply]
  • I still think this should be deleted. The 2000 paper introducing this language has only 16 citations in GS [1], and a couple of those are self-citations. Less than Impromptu (programming environment), which has 27 and is only marginally WP notable in my opinion. FuFoFuEd (talk) 07:07, 15 June 2011 (UTC)[reply]
    • It should be kept The article is improved by adding more reliable and independent sources. 1) Dr Zhou had quitted academy and had worked in industry since 1999 before his NCL paper was published. 2) (Zhou 1996; Zhou 1997), describing a pre-prototype of NCL and scheduling, is only a tiny part of (Zhou 2000). 3) "Ridiculously", (Zhou 1996; Zhou 1997) receives much more attention (16+29 = 45 citations in GS). It is even cited by Prolog's inventor Alain Colmerauer's paper. See comments above. 4) Recent NCL (Zhou 2009) is far more advanced than (Zhou 2000). 5) A book in English has already be finished. And it certainly will help develop the article. 6) Hope this answers your comments. SophiePaul (talk) 17:29, 15 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
  • For comparison, the much more recent ESSENCE language (published at IJCAI 2007 [2]) already has 15 citations on CiteseerX. So NCL with a comparable number over a three times longer interval does not appear that notable in comparison. Another language from the same domain, but slightly older, ECLiPSe has 79 citations [3]. FuFoFuEd (talk) 20:12, 15 June 2011 (UTC)[reply]
  • Keep SophiePaul has added enough arguments. I think it's notable. But please improve. Petterclp (talk) 05:53, 16 June 2011 (UTC) — Petterclp (talk • contribs) has made few or no other edits outside this topic. [reply]
  • Keep This article is notable enough. But improvement might be needed. Logicfan (talk) 07:22, 16 June 2011 (UTC) — Logicfan (talk • contribs) has made few or no other edits outside this topic. [reply]
  • Another argument for the lack of notability of NCL/POEM/MSP is that there's no mention of any of J. Zhou's work in the comprehensive Handbook of constraint programming (2006). FuFoFuEd (talk) 11:09, 16 June 2011 (UTC)[reply]
Answer: That is normal. 1) First I declare to be a user of POEM/NCL. Personally I think, CP is a very vast research domain, so vast that sometimes it is difficult to say what it is exactly. 2) People always tend to think that a programming language should be the result of a big lab. We know that around 1997 J. Zhou was no more than a PhD and he quitted academy in 1999. Nobody was sure what is the future of NCL. At that time, how could that NCL prototype receive enough attention from the academic world? But today, after more than 12 years of industrialization, NCL becomes a much more complete programming language for modeling and solving industrial problems. Everyone on earth knows that industrialization is the most important for a research result. Petterclp (talk) 03:42, 17 June 2011 (UTC) — Petterclp (talk • contribs) has made few or no other edits outside this topic. [reply]
Unfortunately WP:CORP applies to "industrialization", that is commercial products, and this article is not meeting it. The handbook in question does mention the commercial ILOG products for instance, which also originated in academia, so I don't think they are excluding such products by default. FuFoFuEd (talk) 06:12, 17 June 2011 (UTC)[reply]
I am also a user of ILOG (nice products). But from your texts I do not conclude that NCL is not notable. I feel that NCL is different. Petterclp (talk) 08:47, 17 June 2011 (UTC)[reply]
Answer: I use NCL/POEM. I think it does not matter at all if a book on CP did not mention NCL. As far as I understand, NCL stemed from CP. But today's NCL is not exactly within CP, because these articles tell:
  • A very important technique of NCL is its intelligent parser for a context-sensitive grammar. NCL recognizes problem descriptions in "conventional mathematical logic" (coded in TeX).
  • Another point is NCL's algorithmic framework: Mixed Set Programming. At an abstract reasoning level, MSP allows users to formulate problems with a simplified form of first-order logic: quantifiers, Boolean logic, logical functions/predicates, numerical reasoning, set theoretic reasoning, date/time reasoning, etc.
I believe there does not exist another programming language like NCL. Courage, SophiePaul. Logicfan (talk) 12:07, 18 June 2011 (UTC)[reply]
  • Comment. Given the recent appearance of some {{spa}} tags, I think it's worth recalling this text from WP:SPA: "However a user who edits appropriately and makes good points that align with Wikipedia's communal norms, policies and guidelines should have their comment given full weight regardless of any tag." In particular, the original nomination already mentioned that the creator of the page is an SPA; adding a tag to SophiePaul's name at this late stage of the discussion seems redundant. I hope that comments will be evaluated according to their merit. Jowa fan (talk) 13:48, 16 June 2011 (UTC)[reply]
We started with one SPA, now we have three. I also would like this decided on the merits, but that doesn't happen by WP:CANVASSing. Msnicki (talk) 13:53, 16 June 2011 (UTC)[reply]
Attention please: Last night I editted some texts in Wikipedia:Articles for deletion/POEM (software) as a response to FuFoFuEd's challenging comment "no socking fans have shown up to vote keep here". If any problem, don't you think that it was FuFoFuEd who was misleading? I do hope that the debate is kind and serious and there is no pitfall. SophiePaul (talk) 20:50, 16 June 2011 (UTC) — SophiePaul (talk • contribs) has made few or no other edits outside this topic. [reply]
(The comment of FuFoFuEd referred to above is on the page Wikipedia:Articles for deletion/POEM (software). Jowa fan (talk) 00:16, 17 June 2011 (UTC))[reply]
  • Keep. Of course, the article needs to be improved. Dr Zhou has published a book and the article can be certainly further developed. My arguments:
    • "A bit more or few citations in GS" does not necessarily mean more or less notability.
    • Logically speaking, that something is notable with 30 citations does not imply that another with 20 citations is not notable.
    • Reading through the "Related works", I see that Alain Colmerauer, Maarten H. van Emden, Nicolas Beldiceanu, Jaulin el. cited Zhou's work, which is important. I believe that (Martín et al. 2009) and (Flener and Pearson 2004) are reliable and indepedent sources citing NCL; (Li 2009) is a book on an experimental system based on POEM/NCL, another independent and reliable source about POEM/NCL.
    • After reading some of Zhou's papers, I believe NCL is very different from other systems mentioned by FuFoFuEd.
    • Success in industry is very important for a language such as NCL.
    • The conclusion is that NCL is notable. Ortech123 (talk) 06:32, 17 June 2011 (UTC) — Ortech123 (talk • contribs) has made few or no other edits outside this topic. [reply]
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.