torsdag 28 april 2016

Wikidata Wikitree Wikipedia

TBDI was on a session about Linked data in Stockholm and met some people working at Wikimedia in Stockholm and lesson learned.

Wikidata is a container of structured data ex.


English futher down
Jag var på en session om Länkad data i veckan och en ille från Wikimedia var där så vi hade ett intressant prat. Tanken med WIkidata är att på samma sätt som WIkimedia har text skall Wikidata ha strukturerad information.

Idag hämtas datat från exempelvis Wikipedia och en infobox till Wikidata där man kan ställa frågor men det går att hämta information från Wikidata till Wikipedia ex. byter vi Statsminister och det uppdateras i WIkidata så slår det igenom i Wikipedia. Alla kan uppdatera allt lite spännande ;-)

Vad innebär det för Wikitree?

Infobox socken Sverige
{{Infobox socken Sverige
|            namn = 
|            bild = 
|        bildtext = 
|       vapenbild = 
|vapenbeskrivning = 
|        landskap = 
|           härad = 
|          kyrkby = 
|      kyrkbykort = <!-- För de fall där "kyrkby" har ett särskiljningsled
                       såsom "kyrkby = Kyrkorten, Kyrklig kommun", ange 
                       "kyrkbykort = Kyrkorten" för att få en kortare text i infoboxen -->
|             yta = 
|           ytaår =
|          ytanot = 
|      befolkning = 
|   befolkningnot = 
|    befolkningår = 
|       sockenkod = 
|           karta =
<!-- position för för sockenkyrkan --> 
|           lat_d = | lat_m = | lat_s = | lat_NS = N     
|          long_d = | long_m = | long_s = | long_EW = E
}}


Parameter names and descriptions

Name and transliteration

Parameter nameUsageDescription
nameoptionalThis is the usual name in English. If it's not specified, the infobox will use the official_name as a title unless this too is missing, in which case the page name will be used.
official_nameoptionalThe official name in English if different from name
native_nameoptionalName in the local language, if different from name, and if not English. This will display under the name/official name
native_name_langoptionalUse ISO 639-2 code, e.g. "fr" for French. If there is more than one native name, in different languages, enter those names using {{lang}}, instead.
other_nameoptionalFor places with a former or more common name like Bombay or Saigon
settlement_typeoptionalAny type can be entered, such as City, Town, Village, Hamlet, Municipality, Reservation, etc. If set, will be displayed under the names, provided that either name or official_name is filled in. Might also be used as a label for total population/area (defaulting to City), if needed to distinguish from UrbanRural or Metro (if urban, rural or metro figures are not present, the label is Totalunless total_type is se

tisdag 26 april 2016

Templates Wikimedia Family Search

  1. Use Wikimedia pictures on a wiki 
    1. Setting Manual:$wgUseInstantCommons
      [[File:Map of Butler County Ohio Highlighting Monroe City.png|left|thumb]] ==> the picture from Wikimedia is included
    2. Extension:MultimediaViewer

  2. Get A Wikitree and Family Search connect to Wikidata
    1. Request for a property?!?!?
  3. Map Infobox socken to Wikidata
    1. Check status today
    2. How to do it
    3. How do map templates like
      1. Sn-Dm
      2. Sn-NAD
      3. Sn-NAD-ort
      4. Sn-RAÄ
      5. Sn-SOFI
  4. Understand how Infobox socken can be used to feed a Infobox at another wiki

Family Search

  1. Name standard
    Sn-

    Good to have the same as WIkipedia
  2. Set Manual:$wgUseInstantCommons
  3. Lua not installed any plans (Version)

fredag 22 april 2016

How to progress

From threadId=692931732

Areas to discuss/explain 

  1. Family Search Swedish parish pages who is the target group
  2. I have seen 
    1. WikiProject_Sweden
    2. WikiProject:_Swedish_Jurisdiction_Infoboxes
    3. This is the project I am interested in and I have marked Gävleborg County and Västmanland County

  3. How do you track progress and quality

    I have seen that Wikipedia has some matrix of quality and use Maintenance categories how do you work
  4. Using templates inside a wiki I have seen you have done a lot and have some great examples that I played with on my Sandboxe.g.
      1. Navigational template Template:Sweden
      2. The Infobox for Sweden etc...

        Who can create a template and do you have a standard....
  5. External linking best practise?
    1. My experience is using Templates ==> easier to track and easier to change if URL change
      1. Linking restrictions?
        1. Swedish Folklore society? www.hembygd.se
        2. Commercial company Arkiv Digital who "re”scan all Swedish church books
        3. Government sites with good information like maps with parish border Fornsök from Swedish National Heritage Boardvideo I did
    2. Where to place links? When I look on Blekinge County I see the created an “external” page Databases for Blekinge
    3. As you have a “modern” version of Wikitree you have never thought about a navigational box with show/hide?!?!? instead of jumping to another page...
    4. One of the best places about Swedish parishes is Wikipedia they have a lot. 
      1. Is it ok to link to them?
      2. As Family Search is a Wiki and Wikipedia is a Wiki do we have any easy way

        1. Crosslink works using [[Wikipedia:sv:Alfta_socken|Alfta]]
        2. Can we also cross links pictures /maps ex. Alfta church
        3. Or use simplified maps ex. compare on Bollnäs

      3. Coordinates helps a lot finding a place. Wikipedia has done the work and have added coordinates to all parishes churches ex
      4. Koordinater61°20′49″N 16°4′16″Ö
        is done using infobox | lat_d = 61| lat_m = 20| lat_s = 49| lat_NS  = N
        | long_d = 16| long_m = 4| long_s = 16| long_EW = E==> Template Coord is  used
        1. What is best practice inside Family Search for coordinates
        2. My experience is if we “steal” coordinates from Wikipedia that we have a template taking the same arguments
  6. Multi language https://familysearch.org/wiki/sv/Samh%C3%A4lls_Centret
    1. What is the plan? What parts in Swedish….
  7. An area that I believe in is WEB 3.0 or the semantic web ==> today we publish pages in the net and in the future we will link data ==> a resources like Family Search can contribute with  good genealogy information anyone looking into this
    1. Today Wikipedia has excellent information ==> Alfta parish above is Wikidata object Q10405130 that today has no external links 
      but of we look at Picasso Q5593 you can see that a lot of sites has connected to him ==> you can ask for the locations of his (Q5593) works using a
      SPARQL query
      Nathan
       i looked into your profile and I think DNA research will benefit from this open data… 
      We have a Swedish programmer that have done some nice integration between FTDNA and family trees

      see video I did DNAGEN using this tool. If we open up data then more things like this can be done and DNA research would benefit a lot….…



      Another video me "thinking load” about how DNA = FTDNA + Family tree + Sources using templates in a wiki would be magic https://www.youtube.com/watch?v=sAuH6xpsKzA


      The free training
      Introduction to linked data and the Semantic Web from University f Southampton
  8. What is the best place to ask questions that are a bit more technical like
    1. Template design
    2. Cross link 
    3. Bot updates….

Example of a simplified map ==> you add the coordinate and text of Bollnäs and its displayed in the map




Result of a query in SPARQL to find location of Picassos work

fredag 15 april 2016

Sveriges indelning

Länkar


  • Geokod

    GEOKOD är en kodbeteckning för den administrativa indelningen i Sverige 1862-1951. Till skillnad från den mer kända SCB-koden, så är GEOKOD uppbyggd av län/härad/socken, vilket bättre passar släktforskarens syften. Den är helt oersättlig vid bl a registrering av domboksmaterial tack vare uppdelningen i härader. GEOKOD omfattar även fler områden än SCB-koden. GEOKOD är utgiven av Historiska Institutionen vid Uppsala Universitet 1977.
    ISBN 91-554-0552-5.


måndag 11 april 2016

Semantic Web - Linked data

From TED lecture The next WEB Tim Berners-Lee


Suggested sources

  • data.gov.uk

    This began as an initiative back in 2009. Now, data.gov.uk contains more than 9,000 UK government datasets. 
  • item

    data.gov

    This website is the US government's equivalent resource to data.gov.uk. It includes more than 85,000 datasets and includes tools for those interested in playing around with data.
  • item

    G8 Open Data Charter and Technical Annex

    G8 Summit 2013
    Charter signed by G8 leaders on June 17, 2013.
  • item

    Facts Are Sacred

    Simon Rogers
    Faber & Faber, 2013
    Data-related book by the editor of guardian.co.uk/data.
  • item

    Guides to the Open Data Initiative

    Not sure what open data actually is? Wonder how open data might affect you? Curious how to make the business case for open data? Check out these handy guides from the Open Data Institute.
  • item

    World Wide Web Foundation

    I established the World Wide Web Foundation in 2009 to tackle the fundamental obstacles to realizing my vision of an open web that is available, usable and valuable for everyone.
  • item

    W3C

    The World Wide Web Consortium (W3C) is an international group where member organizations, a full-time staff and the public work together to develop web standards. Together with CEO Jeffrey Jaffe, the organization and I are on a mission to lead the web to its full potential.
  • item

    The Web Index

    The Web Index is designed and produced by the World Wide Web Foundation. It is the world’s first measure of the web’s contribution to development and human rights globally in countries across the world. By compiling data across many different dimensions and making it freely available, the Web Index will help deepen and broaden our understanding of how countries can maximize the impact of this powerful tool. It will eventually allow for comparisons of trends over time and the benchmarking of performance across countries, continuously improving our understanding of the web’s value for humanity.
  • item

    Web We Want

    I launched this campaign at a UN gathering in Geneva. Web We Want seeks to engage millions of people across the world to help set a positive agenda for Internet users.
  • item

    Open Data Research Network

    This collaborative initiative is coordinated by the World Wide Web Foundation with support from the International Development Research Centre (IDRC). It exists to connect open data-focused researchers from across the world, bring together information and news relating to research into the implementation and impacts of open data initiatives, and host research projects into open data. The network has a number of projects: "Emerging Impacts of Open Data in Developing Countries” is a multi-country, multi-year study to examine how open data is being used differently around the developing world; the “Open Data Barometer” is a multi-dimensional look at the spread of open government data policy and practice across the world.
  • item

    Open Contracting Data Standard

    The development of a common standard for the disclosure of contracting data is a key pillar of the work to promote disclosure and participation in public contracting — empowering citizens around the world to hold their governments to account for the estimated $9.5 trillion they spend each year through contracts. The Open Contracting Data Standard will be essential to advancing the objective of achieving a new norm in which all public contracting is open. Its development comes as a result of a collaboration between the Open Contracting Partnership and the World Wide Web Foundation, supported by a grant from the Omidyar Network.
  • item

    Long Live the Web: A Call for Continued Open Standards and Neutrality

    Tim Berners-Lee
    Scientific American, December 2010

Books

Wood, David
Sold by: Amazon Export Sales LLC
$35.28

'

DuCharme, Bob
Sold by: Amazon Export Sales LLC
$36.50
Allemang, Dean
Sold by: Amazon Export Sales LLC
$54.95

Training

The free training Introduction to Linked data and the Semantic WEB delivered by Southampton University

Workshops


Example

Below some people in Wikidata with connections to other sources

Douglas Adams has unique Wikidata id Q42 ==> https://www.wikidata.org/wiki/Q42

- Links external 
- - Find a Grave 


Pablo Picasso has unique Wikidata id Q5593 ==> https://www.wikidata.org/wiki/Q5593
with links to



Samuel Adams Q212963 ==> https://www.wikidata.org/wiki/Q212963

Wikipedia 
- Wikidata object Q212963 
Reasonator Q212963
- Wikitree Adams-42 

Samuel Willard Q2221910  https://www.wikidata.org/wiki/Q2221910

Reasonator Q2221910

onsdag 6 april 2016

Wikitree biological vs. Nonbiological relationship

Did a test of the concept that Wikitree also should support social links... see also Wikitree G2G and profile Sjöstrand-29


A father can now be non-biological and appear in Family trees and descendant charts with no explanation compare confirmed by DNA that at least gets an icon.....

It's even more confusing that in a DNA view a descendant appears with the attribute non-biological....
  1. In Wikitree you can have one relation to one father
  2. A relation to a father can today (2016-4) have attributes
    1. confirmed with DNA ==> you have a biological relation
    2. confident
    3. uncertain
    4. non-biological
       
  3.  What is said about using father and say its non-biological is found in the help files
     
    1. In help text the meaning of the father is defined
       
      1. /wiki/Non-traditional_families

        "However, for non-living people, genetic connections should be used when they're known." 

        ==> if it's not stated or the profile has privacy you don't know what is in a field for a living person

        ==> if we don't know the biological parent its ok to use also social links....
          
      2. /wiki/Adoptions_and_multiple_parents

        "For example, if you were adopted you can choose to enter your adoptive or biological parents as your mother and father. If your great grandfather was adopted, you should use his biological parents, if or when they are known.

        ==> hm....but if great grandfather is living you could use non-bio... according to 1)
         
      3. /wiki/Non-Biological "It should be used when you are confident that the connected parent is not the birth parent, but the birth parent is unknown or the family prefers to have an adoptive or step-parent appear on profiles and family trees....

        For deep ancestry it is important to use biological parents when they are known. But for modern, private profiles, it's up to the families. See Adoptions and multiple parents for more information."

        ==> hm....new word that maybe needs a definition deep ancestry is that always non living? Normally I associate that word with Y-DNA testing .....
         
    2. Tests how it looks for an user of Wikitree
       
      1. Family tree view no difference and no explanation between biological and non-biological see Sälgö-Family-Tree-1

        BUT confirmed with DNA has an icon




      2. Descendant view no difference and no explanation between biological and non biological

        see Hagström-Descendants-23




      3. DNA Ancestors view also non Biological relations appear but you have an explanation 

        see Hagström-23/890

        Must be a bug that social links also are listed in a DNA view




      4. On the profile no explanation if the father is father or a step-father....



        See Sjöstrand-29

      5. DNA Descendants view



        OK for DNA tree

        Wrong X-Chromosome also displays non-biological


            
      6. Gedcom Export

        Eva Mona Lisa Sjöstrand with a non biological relationship to Karl Gustaf Hagström is exported as a child

        .....
        1 FAMC @F3@
        .....
        .....
        0 @F3@ FAM --> Wikitree DNA view of family
        1 HUSB @I125@ ---> Karl Gustaf Hagström
        1 WIFE @I126@
        1 CHIL @I3@ --> Eva Mona Lisa Sjöstrand

        Correct is to have 2  FAMC see FAMC-Tag

        GEDCOM 5.5.1 for an adoption use the following structure:
        0 @I57@ INDI
           1 NAME ...
           ...
           1 FAMC @F1@ <pointer to natural parents>
           1 FAMC @F2@ <pointer to adopting parents>
              2 PEDI adopted <[ adopted | birth | foster | sealing ]> 1)
           1 ADOP
              2 FAMC @F2@
                 3 ADOP BOTH <[ BOTH | HUSB | WIFE ] <who adopted>
           ...
        
        
      7. Ancestors list shows nn biological relations with an indication...
        http://www.wikitree.com/treewidget/S%C3%A4lg%C3%B6-1/9
      8. Family Group Sheet no indication if biological or not
        http://www.wikitree.com/treewidget/Hagstr%C3%B6m-23/300
         
      9.  DNA Ancestor Confirmation AID for Magnus Sälgö

        Wrong also displays non-biological relations and explains what tests should be taken

        see Sälgö-1/899

                




 Conclusions/comments Magnus

  1. Using the fields father/mothers both for biological and non-biological just adds confusion
    1. Both from an user perspective
    2. And also the wikitree code seems not built for "reuse" fields in an easy way
  2. Define in a help text that the meaning of a field change over time will just add confusion
  3. A field in a database should have one meaning not changing over time it's bad database design and just adds confusion
  4. From an user perspective not knowing if its a biological relationship you see in a descendant view just adds confusion
  5. In a DNA view display a non biological relationship with an attribute non-biological just adds confusion

Suggestion

Add new fields for social relations and if you should display those relations in descendant views make it clear its a social link not a biological link

See also

  1. Ancestry
    They add an alternative father
  2. My heritage
    They add an alternative father
  3. When Family Means More (or Less) Than Genetics chapter "Multiple Meanings of “Family”

måndag 4 april 2016

Comment Wikitree and hatnotes

Comment to Maryann see G2G at Wikitree
My suggestion is steal what is good and skip the rest....

BUT you must start defining the user requirements you have inside Wikitree... An user on Wikitree doesn't have the same requirements as a wikipedia reader or a family search user.... I started to define what user requirements I found interesting when communicating with people trying to do Swedish genealogy on Wikitree see user story list.

A) This discussion is not about using a template for disambiguation. It's about using a template to format the text added to a category to describe that category.

Magnus: hatnote is friend with disambiguation. Its like footnote and sources... sure you can use sources without footnotes but many people think footnotes makes it easier to read the text.... A hatnote is an easy way to tell someone this is what this page is about BUT there is other pages on xxx that maybe is what you are interested in....

Quote from 
Disambiguation paper above 

Two methods of disambiguating are discussed here:
  • disambiguation links – at the top of an article (hatnotes), that refer/link the reader to other Wiki articles with similar titles or concepts.
  • disambiguation pages – non-article pages that refer/link readers to other Wiki articles.
B) None of the category pages uses a template to format its description.

Magnus: Yes many doesn't and that is stupid... I have also seen they use a lot of templates but not use hatnote so much. They often format it as a hatnote on a page. My humble opinion is that this is as "stupid" as people in Wikitree don't use inline <ref> tag and instead add the number themselves.....

It's not a big deal (look and feel is the same) but with my humble words it's stupid (wikipedia is much more consequent/mature but they started with templates 2004.....) and lesson learned after using templates the last 2 months use it for everything.... maintenance is much much easier....you can never predict what changes you would like to do in the future.... maybe mobile devices requires something else.... I tested last week an Iphone app that is dependent on that a Wiki page had a hatnote format in a special way.... 

C) Some of the category pages also add a link to a main article by using a template. 

Magnus: Yes if I understand you correct I think that is part of the "game" you agree that the Delaware article about the states is "Delaware" main article.....

Maybe that is something for Wikitree..... to state that this is the main article.

(Bad thing inside Wikitree is that you don't have talk pages on categories where things like things like this can be documented/discussed... (they exist in namespace 15 but are not activated on Wikitree))

D)  Categories are used to group subcategories and articles and contains no information about the topic. The linked articles contain information related to the topic

Magnus: I assume you argue about having text on categories or not.... My understanding
  1. The Family Search Wiki is a Wiki about genealogy
  2. The Family Search Wiki is trying to help people do genealogy and have magic articles about how to....
  3. BUT they don't have profiles inside this wiki
so my point is 
  • Wikitree is not Wikipedia 
  • Wikitree is not FamilySearch

    ..... and Wikitree has to find it's own way.
My vote has been and still is use categories to connect information to the profile that makes me understand more/ find new sources.... ==>
  1. Add value to profiles by adding categories
  2. On the category add information = links resources/maps etc....
  3. Make categories easy to use ==> use templates
After spending 3 months getting this started I feel nr 3 is the problem inside Wikitree. Wikitree is a wiki not using templates/see the benefits of using templates.... 

I have tried to get a dialogue with dear Chris and my feeling is that doing wiki with templates is not his priority in life, he feels more it's another headache.... but he doesn't say no.... and as he told me to slow down and Michelle gave up then no one is nearly creating templates any more... and if something is done its just some project graphic with inline css.... 

And maybe Wikitree has a non written policy that priority is to be as easy to use as possible ==> a bio section is just text with a project graphic..... I miss some vision inside Wikitree but I also understand that Chris himself can't "compete" with organisations like FamilySearch/ Wikipedia ....

Using categories to connect a profile to more "knowledge" or find maps with the places or a way to learn more about related subjects is not interesting (for me that is genealogy, adding data from the church book is less interesting...). 

With categories you can connect a profile to knowledge that is related to the profile and get more imnformation about what is on the profile example:

I spent 50 minutes on Family Search and created a sandbox page see User:Salgo60/Sandbox and they have most of the templates I think should be in the toolbox at Wikitree..... so maybe its better to link from Wikitree always to Family Search as they are better on creating genealogy pages ?!?!? They also upgraded last weekend and now have
  • Visual editor ==> end user don't have to play with markup tags. Lesson learned is that I feel the editor is brilliant but people at Wikipedia don't see it as the solution on the problem that "normal" people have problems using a wiki....
  • Modules like Image map ==> clickable maps (Help) ex. AUOHEnglandWI more
  • RenameCategory module ==> easy to change names on categories
  • Better search using CirrusSearch

Example:
Infobox
Template:Infobox_Swedish_Parishes
SidebarTemplate:Sweden-sidebar
Clickable map
Template:Sweden_Image_Map


 .....
Breadcrumbs


                  
                   Navbox that can expand/collapse
                            Template:Sweden



E) Disambiguation pages are not category pages, they are in the main namespace - see Delaware (disambiguation).


Magnus: Yes and that is how Family Search solve it and in my opinion Family Search is not = Wikitree. Wikitree use today a "hatnote" see link and then links on every category to all categories or a link to a main category...?!?!?  is that good/bad. What happened when we get 40 languages.....


F) the template {{Geodis}} is used to disambiguate between articles, not categories, that have the same place-name.


Magnus: Wikitree has problems with name conflicts on categories and todays solution is add a language code and if that is not unique....?

The concept/problem with disambiguation exists in Wikitree and need a solution.... but right now I feel no one focus on that....If a page like Europe (en) is some kind of container for linking to other versions of Europe that could be one way. I think more text is better than just a table with links plus the current design is not working on mobile phones....Wikitree users maybe never access Wikitree from a mobile........ 


G) Template Hatnote: Category-specific templates: none are used to format category descriptions.


Magnus: As said earlier that is as stupid as numbering footnotes by hand and not use <ref tags> or refuse internet for genealogy and start sending postcards ;-) is my humble opinion.... Wikipedia is much more consequent but they started with templates 2004 and as someone said Wikipedia has editors who are young and technical interested which maybe Wikitree doesn't....


H) Category redirects - ... #REDIRECT should not be added to category pages.


Magnus: As said earlier my suggestion is use categories for adding values to profiles, BUT as Wikitree is like Wikipedia before 2004 and are not using templates maybe we should link always from Wikitree categories to FamilySearch articles and then the Redirect is done "over there".... ==> Wikitree has no problems...


  1. A description of the category -- what it should and should not contain.
  2. A note about how you arrived at the category's name

Magnus: That is in Family Search land and maybe Wikitree land should just link to Family Search pages....

Family Search have proven that they are more advanced wiki users and have an organisation for developing pages...


I think Chris has indirect told us that Wikitree prio is not template development or upgrade the Wikipedia software. New development should be done in agreement with a G2G discussions which I feel is Chris and Maryanne asking us to vote on template policies or hat notes....==> I have seen no change the last 6 months... more than some hardcoded tables on Europe (en) using inline CSS and tables

J) A navigation template or navbox is a grouping of links used in multiple related articles to facilitate navigation between those articles.

Magnus: Yes story is that "Normal wiki navigation" using categories is not user friendly. In the Sweden Wikitree category tree last time I looked we had 11 levels ==> 11 click to find the page you are searching after and if we check the usage statistics then very few people access those pages in best case me and Lena...

If its navigation between articles or categories doesnt matter as you have the same problem..... out of the box Wiki categories are not good for navigating more complex structures.....

Example of a navigation template that saves a lot of clicks in a category tree....

Template:History_of_Europe_templates if you have the navigation collapsed as default then its just one line.....

This template expanded


and collapsed ==> one row with a show button to the right


My plan has been to add all category info on a page to a collapsed box ==> takes no space and if people would like to learn more about the content of a Wikitree category then just click on it.....

Status: is a template needs to be developed

Problem Wikitree in General:

We miss a development process in Wikitree and a meeting structure. I have tried with most people to solve this setting up a web hangout.... but this is not the wikitree way today.... 

==> 

We still 4 months later are not moving forward one inch instead people try to develop things under the radar or not at all.... or like me move on to other places....

What is needed for navigational boxes and hide/expand at WIkitree
Creating a navigation template with hide/expand I think we can copy/paste most from Family Search (they dont use LUA as Wikipedia do) and this has been my plan since december 2015.....

Wikitree to do is some templates plus javascript
  1. server based javascript see wikimedia:common.js
    1. See one example jquery.makeCollapsible
  2. server based css to get rid of inline css MediaWiki:Common.css
    1. class navbox...
    2. .... 
  3. create templates
    1. navbox
    2. navbar
    3. Nowrap begin end

Thats it....

But maybe we don't want to have navigational boxes  as its just between articles ;-)
Serious the Europe (en) I feel is a perfect place for a template {{Europe}} that is a navigational box collapsed that can be used on all Europe categories.... what will happen when you should support all languages in Europe (42 languages has more than one million speakers)....

IF
 Wikitree don't start using disambiguation pages which I think is the wiki way of solving problems with namespace.....