Wikipedia:Articles for deletion/Knowledge as a service


 * 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 keep. Consensus appears clear that there is material on which to improve the article because the topic has been covered in a range of sources. If folks would prefer to work on this in draft space, I'm happy to move it, but I see no policy based reason to move it. Star  Mississippi  14:39, 13 May 2022 (UTC)

Knowledge as a service

 * – ( View AfD View log | edits since nomination)

As it stands, it should be immediately apparent from a cursory inspection that this isn't an article. It is an essay. It is Original research. Almost none of it is properly cited to any source. The only sources directly discussing the subject appear to be primary - reports from a conference - which do nothing to demonstrate notability as an independent topic. Other sources cited clearly pre-date the topic, so cannot be discussing it. And as far as I have been able to determine from what few other sources that aren't pure marketing-speak I have been able to locate, 'Knowledge as a service' in as much as it represents anything at all beyond said marketing-speak, consists of the output from knowledge graphs, delivered via the internet. We don't need articles on new combinations of old words used to describe things that already exist, combined in order to make them sound new. If there was anything of merit in this essay that could be merged to the knowledge graphs article, that might be worth considering, but since there isn't, it should simply be deleted. AndyTheGrump (talk) 15:46, 19 April 2022 (UTC)
 * Note: This discussion has been included in the deletion sorting lists for the following topics: Business and Computing. Shellwood (talk) 16:16, 19 April 2022 (UTC)
 * Delete as a service to readers who deserve better than impenetrable marketroid babble. XOR&#39;easter (talk) 16:49, 19 April 2022 (UTC)
 * Delete or maaaybe redirect to Data as a Service, which at least has MORE sources, even if it still reads like marketing babble. My first thought was Content as a Service, based on the title, but it turns out that article is just as bad as this one. PianoDan (talk) 18:06, 19 April 2022 (UTC)
 * Delete as neologism, do not redirect. None of the claims in the lead are supported by sources; the 'Overview' section gestures at the ISWC '19 conference, but no abstracts in that conference actually use this term. &mdash; Charles Stewart (talk) 13:57, 20 April 2022 (UTC)
 * Keep, passes WP:GNG due to significant coverage in multiple, independent, reliable sources.

SailingInABathTub (talk) 01:49, 21 April 2022 (UTC)
 * Are these documents all describing the same thing though? From a quick inspection, I get the distinct impression that they may not be. Instead, it seems that the phrase 'knowledge as a service' is being used in different contexts, to describe very different concepts. One refers smart manufacturing, another to crowdsourcing, another to cloud computing and so on. And these are primary research papers, from a period of over 10 years. If there is an actual well-defined 'knowledge as a service' topic being discussed within academia over this period, one would expect to see overviews of the topic - summaries of the state of research, and at least an attempt to arrive at some sort of clear definition of what 'knowledge as a service' actually is. Articles are about topics, not phrases used in titles of primary research papers. AndyTheGrump (talk) 03:47, 21 April 2022 (UTC)
 * They describe different applications of the same thing.
 * Source 1 says "the KaaS model aims at leveraging the advanced knowledge management methods to yield value by providing knowledge itself as a service, rather than just providing data or information".
 * Source 2, to provide context to KaaS, says, "In contrast to traditional knowledge sharing over the Internet, such as static web pages, knowledge-oriented services (or knowledge services in short) typically aim to provide dynamic, context-aware, and customized information delivery".
 * Source 3, when defining KaaS, says, "Knowledge services are essential infrastructure and key components of the knowledge society, which can be implemented as an IT enabled process that organises knowledge and transforms it into real value".
 * Source 4 says, "The objective of KaaS approach is twofold. First it aims to develop an architecture of knowledge-based platform using cloud computing paradigms focused on the processing of large amount of data (big data approach). Second, it aims to design an architecture of intelligent platform using cloud cognitive services in the context of application in processes of all DIKW levels of knowledge management’s pyramid (machine learning approach)".
 * Source 5, referring to a specific KaaS model, says, "the knowledge service provider acts as an intermediary between the input knowledge and the query responses of the customer, thus creating a context-sensitive environment for knowledge transfer".
 * They fact that the sources cover a large time period is a good thing, it implies a lasting notability. The sources contain overviews of the topic, and "summaries of the state of research" are not required to demonstrate notability. SailingInABathTub (talk) 12:50, 21 April 2022 (UTC)
 * If they are 'describing different applications of the same thing', what is the 'thing'? How, from those sources, are we supposed to create an article telling readers what 'knowledge as a service' is? We already have an essay doing that: original research telling us what the article creator thinks it is. Using the sources you cite to synthesise our own definition of 'knowledge as a service' would put us back where we started. AndyTheGrump (talk) 13:09, 21 April 2022 (UTC)
 * Sure, there is some unsourced rubbish in there but nothing that can't be fixed through the normal editing process. SailingInABathTub (talk) 15:41, 21 April 2022 (UTC)
 * The 'normal editing process' isn't supposed to involves synthesis. Articles need sources which actually tell us what the subject is. Heck, the first sentence of an article lede needs to tell us what the subject is. "Knowledge as a Service is an X..." needs 'X' defined somewhere. Or at least described in a consistent enough manner that we can summarise what sources say about X, rather than merely stating that 'X' is something or other you get via the internet, referred to in several sources. What would you propose should be the first sentence of the lede, and which sources would you base it around? AndyTheGrump (talk) 16:04, 21 April 2022 (UTC)


 * Comment - 's discussion looks promising to me but falls short of persuausive. The third source of the five sources provided builds on a 2010 source with 30 semschol citations A model of knowledge management system for facilitating knowledge as a service (KaaS) in cloud computing environment, which provides an indirect definition, which is that it is a service that provides the same kind of content as a knowledge management system, a well-defined term in library and information science.
 * Having said that, while I could reconstruct a serviceable definition from this source, it's not clear to me that this would clearly be talking about the same thing as this article, nor does 's five excerpts seem to securely anchor the meaning of the term. I agree with 's comment that "The 'normal editing process' isn't supposed to involves synthesis": we can't expect a good result from normal editing if the article isn't unambiguously about a particular thing.  Could SailingInABathTub provide a characterisation of the topic that is suitable for the lead section? &mdash; Charles Stewart (talk) 16:35, 21 April 2022 (UTC)

Relisting comment: To reassess after the edits made during the AFD and the new sources mentioned. Please add new comments below this notice. Thanks, So  Why  08:35, 27 April 2022 (UTC)
 * Comment, as requested by AndyTheGrump and Charles Stewart, I have made a first pass at revising the lede. SailingInABathTub (talk) 12:25, 22 April 2022 (UTC)
 * You cite a 2005 paper as your source. One which states that the authors "introduce and explore a new computing paradigm we call knowledge as a service". The authors are absolutely explicit in this. They are introducing a neologism. One describing a particular 'paradigm' (see fig 1.) which involves extracting data from multiple 'data owners', running it through a 'knowledge extractor' and 'knowledge server', which sends it on to 'knowledge consumers'. A perfectly good model for something, certainly, but no evidence whatsoever that this specific 2005 neologism is what is being described in the sources you cited in your post above. We already know that the phrase 'knowledge as a service' has been used in academic papers to describe various differing concepts and applications. All your new lede does is take one more, and assert that its definition is the valid one. This is WP:OR. AndyTheGrump (talk) 13:38, 22 April 2022 (UTC)
 * "They are introducing a neologism". What is your point? The phrase may have been introduced in 2005 but it has clearly caught on. The other sources that I have added to the article all describe the exact same concept. There is no original research, as everything described in the lede can be found in the sources. You assert that there are multiple definitions of knowledge as a service, but I am yet to see your sources supporting this claim ... SailingInABathTub (talk) 14:41, 22 April 2022 (UTC)
 * Independent notability of a topic has to be demonstrated through sources. Not disproven through sources. The onus is on you. So how about you starting by providing a citation for your claim that the sources you have provided "all describe the exact same concept". It seems obvious enough to me that they don't. Not without stretching 'exact same' to include anything involving sending data over the internet, and calling it 'knowledge'. That isn't a topic, it is jargon. AndyTheGrump (talk) 14:50, 22 April 2022 (UTC)
 * I am satisfied that I have already demonstrated notability. I will await the contributions of other editors. SailingInABathTub (talk) 15:16, 22 April 2022 (UTC)
 * I'm satisfied with this. I've made a small change to SailingInABathTub's definition to make clearer the role that knowledge models play in this 2005 definition. Note that this source has over 50 gschol cites including 39 with that term and two recent citing articles with over 100 gschol cites. For some reason I can't find it on Semantic Scholar, whose count I usually prefer since it allows much less junk in its count, but plainly there are a reasonable number of quality citations of the article.
 * We don't need all the sources to agree; we just need a coherent lead that (i) determines a topic that has sufficient coverage to write an article and (ii) does so clearly enough that editors can grasp clearly what the article is about. That wasn't the case when the article came to AfD but I think it shuld be now. In fact, there is an important difference between the 2005 article SailingInABathTub provided and the 2010 one I did: the 2005 article anchors knowledge in computational knowledge models, while the 2010 article anchors it in the human judgement-based models found in knowledge management. But this is a distinction that we can easily explain to readers and editors in the lead without threatening the coherence of the article. &mdash; Charles Stewart (talk) 20:34, 22 April 2022 (UTC)
 * Keep - there's a substantial literature here and now we have a lead that is anchored in a source, I think we can do a decent job of covering it. &mdash; Charles Stewart (talk) 20:34, 22 April 2022 (UTC)
 *  Relisted to generate a more thorough discussion and clearer consensus.

Relisting comment: Seems to be heading towards delete to me, but let's give this one more chance. Please add new comments below this notice. Thanks, –&#8239;Joe (talk) 11:39, 5 May 2022 (UTC)
 * Comment FWIW, my "delete" opinion remains unchanged. I don't think the new sources mentioned above establish that there is a single, well-defined concept here; stitching them together feels like WP:SYNTH. We write articles about what things are, not what they're called (WP:NOTDICT). XOR&#39;easter (talk) 16:12, 27 April 2022 (UTC)
 * Indeed, there are still self-evident problems with synthesis. We now have a lede citing a 2005 source for a definition, supporting text in the body which asserts KaaS is a new type of "...as a Service" offerings that has been discussed with only nascent examples demonstrated in recent computer (2019) science conferences.... How can something defined in 2005 be 'nascent' in 2019? If the definition in the lede is valid, the (almost entirely unsourced/blatant synthesis) body text is either just plain wrong, or describing something else. AndyTheGrump (talk) 16:44, 27 April 2022 (UTC)
 *  Relisted to generate a more thorough discussion and clearer consensus.


 * Comment - While reflection on the sources has persuaded me that there is a coherent subject here that we can document without violating our policies against creating neologisms or going beyond what the sources say per SYNTH, the current state of the article provides a definition that does not clearly cover all the sources the article cites, and so I am sympathetic to the skeptical delete rationale of . I asked to provide some viable well-sourced lead, and while I saw the problem with the new lead not covering all the sources as is, I regarded having a particular, sourced definition meant that the article would not have the malign effect, currently seen seen in a published 'scholarly' article that cites Wikipedia, of passing off an incoherent concept as KaaS. For me, this moves the article from the category of 'TNT-worthy misinformation' to 'problematic article that can be fixed with normal editing'. There's an endpoint here: revising the lead so that it documents, with adequate sourcing but without violating SYNTH, a broader concept that does either cover all the sources cited or identifies some of the sources as talking about something else. I have some expertise in the field - one of my post-doc posts was in a chair of 'Knowledge Representation and Reasoning' and I've listened to a good number of talks and refereed a good number of articles on knowledge models - and I'm quite confident that the issues are fixable, but I do not expect to have time before the AfD is closed. I'd like the delete voters to consider either withdrawing their delete vote or changing to 'draftify', which is an ATD outcome that puts the article out of mainspace, allowing it to be incubated until it is clearly acceptable. &mdash; Charles Stewart (talk) 03:43, 6 May 2022 (UTC)
 * Can you provide a citation for a source that provides a clear definition of what 'knowledge as a service' is? One which encompasses all significant usage of the term, without being so broad as to lack defined boundaries? Because otherwise, it seems to me that all we are liable to achieve is to go through another round of 'definition by synthesis'. I don't think it is reasonable to draftify without evidence that an article meeting Wikipedia policy can be later created. That needs evidence (rather than just assurances) that there really is a clear topic to create an article on, rather than buzzwords being applied to concepts only linked through the vaguest of commonality, already covered in other articles.
 * As is normally the case, a decision to 'delete' doesn't actually preclude the creation of a new article at a later date, if proper sourcing can be found to demonstrate that there is a coherent single topic, and accordingly I still propose that this article should be deleted, given the failure of edits which have taken place since the AfD started to solve the fundamental issues with the original essay-like state. If anything, they may have made things harder, since they seem to have demonstrated just how disparately the term seems to have been used. Draftifying an article that fails to demonstrate that topic-notability can be met without engaging in synthesis isn't normal procedure, and I really can't see why an exception need be made here. AndyTheGrump (talk) 04:22, 6 May 2022 (UTC)
 * I am not going to make another effort to quote the source in the lead chapter and verse in the course of this AfD. The above argument is the case I have time to make.
 * Draftification is not one of the 'big three' ATDs, but it has its place and experienced closers will recognise when a good case for draftification has been made. Consider WP:NOT bureaucracy: if the article is closed as delete, I will ask the closing admin to draftify; if the admin refuses (some have a policy of not doing this), I will go to WP:REFUND. If that fails as well, I will go to DRV.
 * Hypothetical two-part question: if this did come up at DRV, would you participate in the DRV? If so, what would your argument be? &mdash; Charles Stewart (talk) 05:01, 6 May 2022 (UTC)
 * That would probably depend on the closing statement and decision. And on what had already been said at the review. I would note however that DRV isn't intended to be a rerun of an AfD discussion.
 * Meanwhile, if you want to preserve the article for yourself as a basis for later work, there is nothing preventing you from making a local copy on your own PC. Or at a minimum, at least making a copy of the sources cited, since I think we are all agreed that the article as it stands doesn't properly conform to policy, and needs substantial rewriting to do so. A rewrite starting from scratch from sources which accurately describe the scope of the topic (if such sources can be found) might seem to be the best way to achieve clarity. AndyTheGrump (talk) 12:16, 6 May 2022 (UTC)
 * To be clear, I would only go to DRV if draftification was blocked as an option. I prefer to work collaboratively on rescuing articles, in draft space, using the WP:AFCH tools where appropriate. I see no case made for blocking this route. &mdash; Charles Stewart (talk) 14:35, 9 May 2022 (UTC)
 * To be clear, I would only go to DRV if draftification was blocked as an option. I prefer to work collaboratively on rescuing articles, in draft space, using the WP:AFCH tools where appropriate. I see no case made for blocking this route. &mdash; Charles Stewart (talk) 14:35, 9 May 2022 (UTC)


 * Keep. Definitely a "there" there. Hyperbolick (talk) 02:47, 9 May 2022 (UTC)
 * Keep. Quick Google Scholar search shows that this concept definitely meets GNG (t &#183; c)  buidhe  09:43, 12 May 2022 (UTC)


 * 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.