Archives |
---|
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 |
|
Welcome to the external links noticeboard | ||||||||
---|---|---|---|---|---|---|---|---|
|
||||||||
Additional notes:
|
||||||||
To start a new request, enter a report title (section header) below:
|
Indicators |
---|
Defer discussion: |
Defer to RS/N |
Defer to WPSPAM |
Defer to XLinkBot |
Defer to Local blacklist |
Defer to Abuse filter |
Contents
Drexel Dragons men's lacrosse
Drexel Dragons men's lacrosse ( | talk | history | protect | delete | links | watch | logs | views)
The external link for "Drexel Lacrosse A History of the Heart - William Thayer" seems to be setup so that the pdf file it's linked to downloads to your computer, instead of opening up in your web browser. Not sure if this kind of thing is acceptable per WP:ELNO. It could just be the way my computer is set up, but even so not sure if something needs to be added to the link's description so that the reader is aware that the file may start to download. Is there a way to tweak the url address so that clicking on the link does not automatically start to download the file? -- Marchjuly (talk) 00:17, 16 February 2016 (UTC)
- Update: The link in question was removed by another editor for another reason, so this problem appears to have been resolved. -- Marchjuly (talk) 22:24, 26 February 2016 (UTC)
Links to some sort of thesis about Islamic eschatological figures
Are [1] and [2] acceptable external links for the articles Mahdi and Masih ad-Dajjal? They would most certainly not pass as reliable sources, being self-published and all that (with a dash of WP:COI), but the external links guidelines do not seem to care about reliability. Pinging FreeatlastChitchat and Elzbenz, who have also dealt with Abdullahfaruqibnibrahimy's links, but I'd like some 3rd party opinions too. Thanks! - HyperGaruda (talk) 19:46, 21 February 2016 (UTC)
- I deemed that the author was too obscure, based on his Google Plus profile and the fact that the paper isn't published , but just uploaded on Google Drive. My decision was somewhat rash though, as I am 100% ignorant of the subject matter, and didn't pay any attention to the text in my decision. It was just that the account is the name of the author of the papers, so there was a conflict of interest in adding them, and I had never seen obscure unverified research papers linked in External links before. They should be removed I think, unless someone else vouches for them that isn't the author. Elzbenz (talk) 01:32, 22 February 2016 (UTC)
- Appears to be a research paper submitted to some university. I am not sure what the policy is on those. WP:RS is only for sourcing content, I am sure the policy for links will be a ....... a bit lenient. My revert was more on the lines of COI. FreeatlastChitchat (talk) 03:18, 22 February 2016 (UTC)
- Said user has been at it again, this time adding a link to a youtube video on his personal channel. By now, I think we can safely call this WP:BOOKSPAM and WP:LINKSPAM. I will revert said user's edits. - HyperGaruda (talk) 16:46, 23 February 2016 (UTC)
- Appears to be a research paper submitted to some university. I am not sure what the policy is on those. WP:RS is only for sourcing content, I am sure the policy for links will be a ....... a bit lenient. My revert was more on the lines of COI. FreeatlastChitchat (talk) 03:18, 22 February 2016 (UTC)
Links added to Ark of the Covenant
The following has been repeatedly added to Ark of the Covenant by two editors with few or no other edits:
- Stone, Adrien E, The Ark of the Covenant - Decoding Exodus Ark of the Covenant - Do it Yourself Project. From Secretofthevine.com the online Spiritual Curriculum.
My instinct is that this is nothing more than spam. Others' views would be appreciated.- MrX 03:31, 22 February 2016 (UTC)
- Definitely. Basically unknown article, clearly self-published. But then I'm on of the editors removing it when added. Doug Weller talk 19:23, 26 February 2016 (UTC)
Use of external links in lists of datasets
Hello everyone,
I was referred here by another Wikipedia user concerning the usage of external links on pages like List of datasets for machine learning research and Comparison of datasets in machine learning. First, some background:
Machine learning is a field where automated methods are used to find patterns in data. It has been argued [1] that the availability of high-quality training datasets is actually far more important than the availability of new machine learning algorithms. It appears that in many/most cases, great technological leaps in machine learning performance come when a new high-quality dataset is made available, not a new algorithm (some examples of this given in [1]).
For these reasons, Wikipedia should have some record of the groundbreaking datasets that have helped to significantly advance machine learning research. Since the vast majority of individual datasets do not have their own articles (and despite their importance in the field, probably shouldn't), articles like List of datasets for machine learning research and Comparison of datasets in machine learning can serve to aggregate the most important among them.
To maximize the utility of these pages, they contain external links to the webpages where the datasets mentioned can actually be downloaded and used. In both of the example articles given, these links are placed in the rows of the tables. I wanted to open up a discussion here about the usage of these links and their compatibility with WP:EL. And, if they are not compatible, begin the process of discussing an exception. I see in WP:EL that "lists themselves should not be composed of external links." However, it also says "This section does not apply if the external link is serving as a citation for a stand-alone list entry that otherwise meets that list's inclusion criteria."
In closing, I believe pages like these will become great resources to the machine learning community as it continues to develop. I look forward to discussing with everyone. DATAKEEPER ✉ 18:49, 26 February 2016 (UTC)
- ^ a b Wissner-Gross, A. "Datasets Over Algorithms". Edge.com. Retrieved 8 January 2016.
- WP:EL is very clear on this. There are no external links for the list entries in a standalone list, much less being inside the body of the article. These links are to datasets and are not citations for inclusion in the lists. From WP:EL: "These lists are primarily intended as providing direct information and internal navigation, not a directory of sites on the web." List of datasets for machine learning research with external links is solely a way to provide links outside of Wikipedia and not for internal navigation. I see no need for an exemption from how Wikipedia treats external links for this article. The best way to have an external link to one of these datasets is to add it as an Official Website link at the bottom of the article about that dataset. But, none of these datasets have an article to add an external link, which further casts doubt on if this standalone list should be included in Wikipedia. Please remove these links immediately or revert back to the edit that removed them previously. Thanks, Stesmo (talk) 03:15, 27 February 2016 (UTC)
-
- @Stesmo: - Hello again, and thank you for clearly stating your stance on this. For everyone's information, I came to this page and posted this question at the request of Stesmo, which can be found on their talk page. Since Stesmo was the one that referred me here, I was hoping to get an impartial 3rd party opinion on the matter before laying it to rest. Thanks again everybody. DATAKEEPER ✉ 18:56, 29 February 2016 (UTC)
-
-
- @Datakeeper:I don't agree with Stesmo's opinion (on his talk page) that these links cannot be added to the references. Most of the linked webpages provide all the info listed in the table, so they are a reference for that info. Removing them would leave an article about an IT topic with only 3 of 180 references available online. If they can't be in the table, make them references or add them to the references. You may want to get additional opinions before making such changes (I trust you can write a script to make such changes, would be rather tedious doing it manually). Prevalence 04:11, 2 March 2016 (UTC)
-
@Datakeeper, Prevalence, and Stesmo: - First, the way they are currently linked is inappropriate, at best they should be turned into references, or in a far-to-the-right column with 'link to dataset'. And for the latter I would still disagree that these links serve Wikipedia's purpose. They are in a way nothing different from direct links to the websites of restaurants in New York on the article 'restaurants in New York'. And both of those fail WP:NOTYELLOW.
Now, regarding the first entry in the first table on List of datasets for machine learning research there is an external link to FERET. That is a notable entry in itself, as witnessed by our internal article FERET. The external link there should hence be converted to the internal link to the internal page - and on the article FERET a link to the dataset (or better, the 'homepage' of the dataset) IS appropriate (and that is there). It does not belong in the list. I haven't checked any other cases, but I presume that there are many which do have internal links possible.
So shortly, all external links in the first column should go, and should all turn into internal Wikilinks. For items that are not notable in Wikipedia terms (hence, are redlinks) sufficient references should show some notability (as in, reliable independent sources must at least have 'noticed' the database), and could contain a reference to the 'homepage' of the dataset. For items that fail that notability test, they should simply be removed from the list (what we see on other similar list-pages is that people just add their self-crafted example to such lists - anyone can make a database of something, post it online and link it here, that is not enough for being mentioned in Wikipedia). --Dirk Beetstra T C 05:30, 2 March 2016 (UTC)
- @Beetstra, Prevalence, and Stesmo: Thanks everyone for the comments and thoughts. I know everyone wants the best for Wikipedia, and I appreciate any time spent on this. Beetstra, I wanted to address a few of your comments and concerns. (1) A small number of the datasets listed have their own articles — these I will link in. Others are noteworthy and included due to their repeated mention in peer-reviewed machine learning research. FERET is sort of a special case because it is extremely famous. I've made sure that every dataset included on the page has at least one independent, peer-reviewed reference proving its notability. (2) I'm fine removing external links from the first column, but this will leave a large number of the datasets without internal links.
- How about I write a script to move all of the external links to the references that are already there for each dataset? Simply appending a "link to this dataset" or "dataset homepage" to each item? Would this be a solution everyone can accept? Thanks again. DATAKEEPER ✉ 20:32, 2 March 2016 (UTC)
As explained above, the links violate WP:EL and WP:NOTYELLOW. Changing the links to references only makes the problem worse by pretending that the links are references when in fact they are there to serve as a directory. The solution is to write the article first, for each notable entry, then link to the relevant article in the list. --Ronz (talk) 23:53, 2 March 2016 (UTC)
@Datakeeper: - I think it would be good to see what happens when you try to link the proper articles. I do hope that the references for the then redlinked ones do show reasonable notability over being just mentioned as 'yet another dataset'. As I said above, everyone can create databases of whatever, that does not mean that it should be in the list. I think that the bar should be that the dataset is being used in some reasonably visible or important way. --Dirk Beetstra T C 06:31, 3 March 2016 (UTC)
@Ronz: Thanks for the input. As was discussed above I'm fine moving the links out of the table altogether. I suggested what others suggested, moving the links to the end of each reference as a link to "dataset homepage." This could easily be done with a script. I do not plan to change the links in the first column of the article to internal Wikipedia references, because most datasets (even those used in a great deal of research) do not really warrant their own articles and so they don't exist. @Beetstra: this applies to your comments too. I do want to address your concern that "everyone can create databases of whatever." This is definitely true! I wanted to assure you that, as evidenced by the references, every single item in this article is not only of value to machine learning research, but has been referenced in one or more academic journal publications on the topic. This list is not just an indiscriminate group of datasets from various places.
So, to resolve this and avoid violation of WP:NOTYELLOW & WP:EL, there are two options I see:
- Remove all external links.
- Move external links from table to tail end of each reference as a "link to the dataset homepage" as was previously preposed.
Because it could be argued from WP:EL that these links are indeed serving as citations, I think the second option is the better of the two. I've heard opinions on both sides and I think the second one will result in a better and more useful article (while abiding by WP:EL) as it grows as a resource for the machine learning community. How about I try it out? I'm fine personally making these edits. Thanks everyone. DATAKEEPER ✉ 19:02, 3 March 2016 (UTC)
- Any link that isn't clearly a reference would be blatant linkspam/refspam. Assuming that all links would be appropriate as references ignores our polices and guidelines. This is a common problem Wikipedia has with spammers. Let's be sure to steer clear of what could be seen as an attempt to circumvent Wikipedia's policies and guidelines in in order to promote non-notable and unencyclopedic information.
- You're overlooking the obvious solution: delete the article as inherently unencyclopdic per WP:NOT. --Ronz (talk) 19:25, 3 March 2016 (UTC)
- @Ronz: Each link clearly is a reference. I do not assume this. Additionally, I assure you that there is no hidden agenda to "circumvent Wikipedia's policies." It is for compliance with these policies that I opened up the discussion here. I appreciate your concern for the quality of Wikipedia and adherence to its policies - I share this concern with you and so have made sure to open up a discussion here when the question of external links was raised on the article under discussion. Also, I strongly disagree with the notion that the article should be deleted. This discussion is centered around external links, not the deletion of the article as a whole. Thank you for stating your opinion on that matter. If the consensus seems to be that the external links should go, I'll be glad to remove them. DATAKEEPER ✉ 19:45, 3 March 2016 (UTC)
- I looked at the first link in List of datasets for machine learning research, and it doesn't look like a reference to me. Is it an exception or the rule?
- Lists of non-notable entries are problematic on a number of levels, comparisons worse. They need to meet WP:N and WP:NOT to start. --Ronz (talk) 20:40, 3 March 2016 (UTC)
- As discussed above FERET is actually one of the very few datasets with its own article. As for the link being a reference: the given link verifies the contents of several of the columns in that row. DATAKEEPER ✉ 22:12, 3 March 2016 (UTC)
- @Ronz: Each link clearly is a reference. I do not assume this. Additionally, I assure you that there is no hidden agenda to "circumvent Wikipedia's policies." It is for compliance with these policies that I opened up the discussion here. I appreciate your concern for the quality of Wikipedia and adherence to its policies - I share this concern with you and so have made sure to open up a discussion here when the question of external links was raised on the article under discussion. Also, I strongly disagree with the notion that the article should be deleted. This discussion is centered around external links, not the deletion of the article as a whole. Thank you for stating your opinion on that matter. If the consensus seems to be that the external links should go, I'll be glad to remove them. DATAKEEPER ✉ 19:45, 3 March 2016 (UTC)
As I said, I'm fine removing the links if that is the consensus. It seems to be the majority. I'd be happy to resolve this. DATAKEEPER ✉ 22:12, 3 March 2016 (UTC)
- @Datakeeper and Ronz:. I have now looked at the second item in the list (that the first one is notable is already established, we have an article on it). We are talking there about Pose, Illumination, and Expression (PIE), which is a database created by Ralph Gross, with the help of PhD student Terence Sim and under supervision of Simon Baker. You reference that to "Sim, Terence, Simon Baker, and Maan Bsat. "The CMU pose, illumination, and expression (PIE) database." Automatic Face and Gesture Recognition, 2002. Proceedings. Fifth IEEE International Conference on. IEEE, 2002.", which is obviously not an independent reference. The next 4 references clearly state the same names as people who created the database in question. I state above "I do hope that the references for the then redlinked ones do show reasonable notability over being just mentioned as 'yet another dataset'. As I said above, everyone can create databases of whatever, that does not mean that it should be in the list. I think that the bar should be that the dataset is being used in some reasonably visible or important way." For the first handful of items below FERET you have not been able to show that (you merely show existence, which is not enough), nor do I think that these databases are having their own Wikipedia article. I will tag the article with {{primarysources}} and {{notability}} - and I do hope that you will resolve the issues. --Dirk Beetstra T C 03:41, 6 March 2016 (UTC)
-
- @Beetstra: Thanks for looking the article over. I appreciate the keen eye and feedback. The topic of that article - datasets used in machine learning research - is a notable one - this is established in the introduction. Data plays a key role in shaping the discoveries in this field, and so famous datasets are ones that contribute to significant discoveries in the field. As machine learning and artificial intelligence continue to advance, the availability of high-quality datasets for research will be one of the key drivers of innovation. Due to the unarguable importance of groudbreaking datasets to this field as evidenced by the literature, the notability of this article is established. High quality datasets have significant coverage in reliable (peer-reviewed and other), independent sources.
-
- The inclusion criteria for this list is: included datasets have been used in academic peer-reviewed literature or books. This is to prevent inclusion of various, not-noteworthy "yet another dataset[s]." For them to make it into literature and be cited, they have to be significant. The purpose of the reference column in the article is to show that it is worthy of inclusion in the list. As for your example of "The CMU pose, illumination, and expression (PIE) database," a Google Scholar search (here) will shows that the reference included has been cited 2429 times in academic literature, which makes it extremely significant. The reason I included the primary source is so people can find later find the dataset, but perhaps this would be better to be left Google searching. I'll make sure to go through the sources and add secondary sources to the places where primary sources are used. Since this discussion no longer centers around external links, let's take any further discussion to the talk page of the article. Thanks again everyone for the great discussion. DATAKEEPER ✉ 05:40, 6 March 2016 (UTC)
-
-
- @Datakeeper: and that is exactly where the problem lies - 'included datasets have been used in academic peer-reviewed literature or books' alone is not enough. Notability means that independent parties have used the database for some independent research or real-life use. The references that are there are not showing that, and even having 2429 articles referring to that paper is not enough. If all of those are 'look, there are more databases with faces' then that is not an independent use of the information, let alone that that 'use' is notable enough to be included in Wikipedia. Moreover, they may be referring to the techniques used in the articles, not to the database itself. I will grant you, with 2429 references referring to the article there are prone to be a couple which show independent importance, but the current list of references does not show that at all - and hence you are not showing notability of the individual items.
- I am not questioning the notability of the subject, but a list article is not a proper place to show notability of the subject. Machine learning is the place for that. --Dirk Beetstra T C 05:57, 6 March 2016 (UTC)
-
-
-
-
- @Beetstra: Thanks for the clarification, I will cite secondary, independent sources that use the datasets for their real-life research. Continuing with our example of the second dataset, you will see that many of those sources are real-life applications of the dataset to train machine learning algorithms to perform tasks. Additionally, this article is distinct from Machine Learning in that it is a list of the most important datasets available in the field. As the coverage of machine learning and its sub-topics increases on Wikipedia, I expect many of these datasets could have their own articles. I'll be sure to put in some edits soon to address your concerns. DATAKEEPER ✉ 06:42, 6 March 2016 (UTC)
-
-
-
-
-
-
- @Datakeeper: What I meant was, that the base subject is notable, and that the databases are of importance is quite clear. However, if that can be covered by 3 or 4 databases that show significant real life notability then there is no need for the list-article (that can be part of the base subject). There can be 2429 references to the paper of #2 .. but if none of those scientific papers (or mainstream reports outside of this count) show that that specific database is currently in use in some real-life form, then the notability may not be sufficient for Wikipedia to include them in the list. Being used in the 'grander scheme of things' does not necessarily make the parts notable either. We'll see what happens to the 200 items in the list now. --Dirk Beetstra T C 07:59, 6 March 2016 (UTC)
-
-
-
List of company registers
Well, this is my first time behind the curtain here at Wikipedia (long-time user and financial contributor) so not sure how to go about this. Someone, and I won't mention whom because apparently I have to notify them with some string of characters I don't understand; someone keeps removing the links on this very useful page that lists all government registry websites and when you click on the name it takes you there. Apparently there are some pedantic reasons why the links shouldn't be there but it kind of makes the article useful to people who use it. I read the WP:EL thingy and I really think there is enough wiggle room to allow the links to stay. Especially considering that every time someone takes out the links someone puts them back in. Now I wonder is it the people who use the article that are taking them out or is it someone who really has no use at all for an article about company registers? Its locked now so I'm here trying to get it unlocked and the links put back in. — Preceding unsigned comment added by The Westernerer (talk • contribs) 19:13, 9 March 2016 (UTC)
- Wikipedia is not a directory. The links don't belong.
- Given Company registers doesn't exist, why does this list? --Ronz (talk) 21:23, 9 March 2016 (UTC)
- Well, how about instead of deleting nearly 40,000 carefully gathered and useful links they get moved to External Links? — Preceding unsigned comment added by The Westernerer (talk • contribs) 21:49, 9 March 2016 (UTC) The Westernerer (talk) 21:51, 9 March 2016 (UTC)