We are seeing a growing number of person-pages with names that match Wikipedia pages except for details such as dates. There will be more.
Several years ago the consensus was that each such page should have a redirect from a Familypedia page that matched the Wikipedia name exactly. Two reasons:
- SEO (Search Engine Optimization) - people searching the web are more likely to find our page if it (or its redirect) has the name they search under
- Internal help and minimizing of duplication - when an editor wants to create a page for a WP figure, the presence of a page named exactly as on WP will minimize the chance that the editor may unwittingly create a page in our style but with a slightly different name and do a lot of unnecessary work because there is no warning of duplication
The resulting pages such as Barack Obama are a little at variance with our standard hndis pages, such as Mary Brown. Our preferred solution is to create a page such as Barack Obama (disambiguation), in the same style as Wikipedia. Sometimes that does not match a WP disambiguation page, a common reason being that WP has only one person of that name whereas we may have several. It is not hard to adapt the standard template {{SPN}} - "Similar person name" (which redirects to {{SimilarPerson}} ), to cater for a " (disambiguation)" page, but it is irritatingly fiddly to have to do so. The proliferation of such pages suggests to me that a simpler solution is desirable and may be available. Could our experts devise a variation to {{SPN}} that ignores the "(disambiguation)"? One approach could be to change the standard to "/disambiguation" so that the coding could merely replace {{PAGENAME}} with {{BASEPAGEAME}}. I have checked the coding to see whether that would completely solve it. I think changing to BASEPAGENAME in that template would solve it; one included template contains {{PAGENAME}}: ((T|Similar person names}}, but in harmless contexts; {{Same name}}, the SMW insert, does not contain PAGENAME.
One problem with making the subpage the new standard would be the need to recode the existing pages a bit. Bot job.
Thoughts on renaming and recoding?
-- Robin Patterson (Talk) 23:42, June 16, 2016 (UTC)
The basic principle of Familypedia is that any person can be the object of an article, while Wikipedia has rules on the notability of each person, rules which are not applied consistently in different languages editions. This inevitably will mean that Familypedia will have more persons identified by the same name than Wikipedia. It would be incorrect to have redirects for a name from Wikipedia, (whithout dates) to the article of the same person in Familypedia with dates. It would be incorrect and unfair because it would imply that a person with a certain name who has as article on Wikipedia is entitles to a redirect, while all other persons with the same name would not have the redirect. Therefore the only correct solution is to use the disambiguation page which would include all the persons with the same name.
What should be taken into account is that Familypedia is designed to be a genealogy database. Therefore it would normally not be used by people who are interested in celebrities. Person who want to find out something about Barak Obama, president of the United States will normally not look into the Familypedia page. Therefore designing a system only for celebrities is not what Familypedia needs.
The simplest way would be to use an automatic system to create disambiguation pages for all the person names in Familypedia. Until then, wherever needed, they can be created manually. Though I am not sure that they are needed. For instance, let us assume that somebody is looking for József Somssich de Saárd. Familypedia has 5 persons with this name. Assuming somebody does not know which one he is looking for (or simply does not know when the person he is looking for was born) and just types the name, without any dates, the system will automatically list all the five persons called József Somssich de Saárd with their birth and death dates - exactly what a disambiguation page would do. Thus the search engine of Familypedia solves the problem, even without a disambiguation page.
Actually, if you look for Barak Obama, the same Familypedia search engine will show that Familypedia has two persons with this name, but also includes some tree, descendants, ancestry, info pages for the person with this name and also some articles with other names who mention persons called Barak Obama. Would a disambiguation page provide any more information to the person who is searching? Would another more sophisticated system provide anything more?
If we are looking at the results of the SEO results on the internet, it is easy to assess that by using the search engine of Google or other browsers, finding the Familypedia article for a celebrity does hardly provide satisfactory results. Searching on Google for Barak Obama or for Barak Obama (1961) does not improve the search results. None of the searches enables the person who is searching to find the Familypedia article, though you could identify the article in searching for "Barak Obama (1961)". Neither the disambiguation page nor any other solution would improve the chance of the internet search identifying the Familypedia article.
Therefore I suggest that the general solution for Familypedia should be to rely on the search engine which provides satisfactory results. If we are thinking of another solution, let us first identify what the Familypedia search engine does not provide and what we want to improve.
Afil (talk) 01:38, June 17, 2016 (UTC)
- I like the the present standard of redirecting the Wikipedia names, since many of the pages we copy from Wikipedia have links to people. A recent couple of mine are Droitwich Spa and Earl of Clarendon. This has nothing to do with search engines! Thurstan (talk) 03:03, June 17, 2016 (UTC)
- Thurstan's point is well worth remembering. Pages I have created in the last few days from Wikipedia mostly have in their narratives several links to other WP (and ultimately FP) person-pages. Andrei's proposal would require us to edit every such article so as to add "(disambiguation)" to all of those inline references. Sticking with Wikipedia page names except in a few instances (e.g. populated places) saves thousands of hours of work and lets us use a greater proportion of our time for genealogical writing.
- I agree in principle, but not in practice, with much of what Andrei said above. I think he overdoes references to celebrities (a term that really doesn't apply to the writers and scientists and engineers who pepper our WP people category) and overestimates the value of the wikia search engine, which can often give far too many hits to be useful. I've just done a search for "john cooper" (the second name I randomly thought of, the first one having been already the subject of a hndis page); the first five hits are women and the sixth (of over 40,000!) is John Alden, who was hired as a cooper. Not "satisfactory". Even the few lines of text you get are less useful than the SMW-based columns on a hndis page.
- I don't think people searching for someone who shares a name with WP's choice for the plain name would be noticeably put out by finding that the person of apparent less notability was on a disambiguation page, and they would not be likely to worry or even think about "entitlement"; if they were looking for the favoured one they would be pleased to get a redirect straight to it rather than having to look through a disambig page; if they were looking for another, they would find it from the disambig page just as fast under our current system as from Andrei's ideal system - or maybe faster, because under Andrei's system, as suggested above, we might not have had time to create the disambig page.
- A Yahoo search for "barack obama genealogy" puts one of our pages in third place behind two other genealogy sites and ahead of Wikipedia. But that's not a good example either way, because the search engine grabbed the "barack obama" from a page that added "(1961) immediately after. Better examples would involve our use of middle names. But who knows what a search engine really looks for? I've just asked Google for "James Arbuthnot" and got a couple of dozen variously-named hits for the British MP, with our page "James Norwich Arbuthnot (1952)" in 14th place; well ahead of thePeerage's "James Norwich Arbuthnot - Person Page"; with no sign of our "James Arbuthnot", which SEO theory suggests should have been fairly high on the list because of the proximity of the words.
- -- Robin Patterson (Talk) 12:25, June 17, 2016 (UTC)
- There is nothing we can do about the Google or Yahoo search engine. But would it not be possible to improve our own Wikia search. For instance: the search engine does not search only for the name in the title, but also to the articles where this name is mentioned. So in we search for John Doe, the search engine will not present only the articles about the various John Does, but also those where John Doe is mentioned, and even those where only John or only Doe is mentioned. This may help some searches. But would it not be possible to simply sort the results of the search, so that the articles where John Doe appears in the title, either just as John Doe, or with John Doe with dates and John Doe subpages (such as tree, descendants etc.) and just afterwards the articles which contain references to John Doe. The results which have only John or only Doe should be listed last.
- The rationale of this is that any search engine tries to satisfy the needs of a greater number of people searching for information. The needs of the general population of searchers, which google, yahoo or bing try to satisfy, is obviously different for those searching Familypedia, where the searches are more genealogy or person oriented. Therefore a general search engine is not best suitable for our needs. The question therefore arises, who is in charge of the Wikia search engine and to what extent can the Wikia team adapt the search engine to the specific needs of a certain field such as Familypedia? Afil (talk) 00:18, June 18, 2016 (UTC)
- Wikia is in charge of the Wikia search engine. I have more than once complained about it. It used to show first all pages that had the search term in their titles, followed by pages that had it in text. Not any more, I suspect; and I see no future in trying to get Wikia to adapt its search for just one of its 360,000 sites. I think Google treats us better. I believe we must accept Wikia's limitations and improve our FP-specific search features. -- Robin Patterson (Talk) 04:09, February 22, 2017 (UTC)