Friday, January 22, 2016

#Wikipedia - Nbr of statements on #Wikidata

Edo de Roo produced some stats showing the number of statements on items that have a Wikipedia article in Dutch.

I like his approach; it shows how well Wikidata might serve a Wikipedia. It may infer how well subjects that relate to the Netherlands are supported on other Wikipedias as well. There is more about subjects related to the Netherlands than what you find only on the Dutch Wikipedia but that is a bit more complicated.

When you compare these numbers to Magnus's stats, it is not bad at all. Most items for all of of Wikidata have 1 statement (4,474,865 23.22%) and for the Dutch it is 6 statements (36078).

Obviously there is room for growth.

Thursday, January 21, 2016

#Wikimedia - Perfection is the enemy of the good

I was wrong to tell people that Mr Anthony needs an article. I was wrong because the item I added on Wikidata was not perfect and, indeed it is not. I was wrong because among the awards Mr Anthony received was the "President's Award for Distinguished Federal Civilian Service".  It says so on the website of the Boston University..

There are several approaches to such criticism. It is true, obviously. I made the point that Mr Anthony received awards to do with psychiatry; he is very important in this field and deserves at least some of our attention. Mr Anthony received indeed the President's award. It is awarded to only five people a year.. It makes him notable in the biggers scale of things. It could be a reason for Wikipedians to take note of him and the other people who were celebrated in this way.

I am proud that I make mistakes; it proves at least that I do something. I think that this is a worthwhile thought. When what is done is not perfect in the eyes of others, so be it. It is all too easy to find fault. When many people make an attempt to do good, it is wonderful. It is how Wikipedia became what it is. It is not perfect and by finding fault at what others do, attention is diverted from what makes a project good, even better.

The aim of the Wikimedia foundation is "to share in the sum of all knowledge". Arguably our projects serve what all our editors put in there. Arguably Mr Voltaire already knew that perfection is the enemy of the good. My sense is that the arguments around the Wikimedia Foundation, its software, its communities have lost much of its validity. It is a bit like with psychiatry, for many psychiatry is business as usual: rehashing the old arguments, iterating on the method of operandi and applying the same stigmas. It takes people like Mr Anthony to bring hope, to tell people that there is room for recovery, that it is important to (re)connect to the values that are important.

There is hope for Wikimedia and it is not in rehashing time and again what went wrong. It is in what went right and what it takes to make it go right again and again. It is in reconnecting to important values like "be bold" and in recognising the work people do.

Wednesday, January 20, 2016

#Wikimedia - What I wish was strategic

The Wikimedia Foundation asks people to consider its future. This is part of a strategic reevaluation of what it does. The aim is to renew the vision so that in the future our readers will be optimally served.

  • Our objective is to "share in the sum of all knowledge". This should stand but practically we should "share in the sum of all available knowledge"
  • We should research how we can improve our reach for all our projects. 
  • Wikisource should get an additional front end that is mostly about presenting books that are finished to the public. This will likely exponentially increase the relevance of Wikisource
  • Wikidata could be central to many processes in other projects. The problem is that only in Germany development work is done. It is why opportunities like replacing wiki links and red links are not considered to improve quality.
  • We need to map where our projects are weak. There are many relevant subjects that are underserved. Gender diversity has a momentum of its own nowadays so let us focus more on the subjects where we are really weak.
  • The Wikimedia Foundation is more than Wikipedia. The singular focus it gets diminishes our other projects and as a result we do not realise what we aim to achieve; "share in the sum of all knowledge".

Sunday, January 17, 2016

#Wikipedia - 15 years

A long time ago in a country far away, they experimented with the creation of new content for an encyclopaedia because their current model proved broken. They allowed people just to write and the idea was that when it was 'good enough', it would be used in the encyclopaedia.

The new mouse trap, using a 'wiki' approach, proved effective, so much so that it easily outpaced the old encyclopaedia and replaced it. The new model became Wikipedia and people were amazed. It should not work but it did. It embodied values that were very much outside the established models of science and publishing. It broke those models and ever since, science and publishing are trying to catch up.

As it did not fit in the existing paradigms, ever since people have been eager to insist on implementing their old hat thinking because .. whatever. Wikipedia grew. People added content and some of it 'did not fit'. It is why new projects like Wiktionary, Wikinews and Wikisource started. They have their own history and community. After their effort was sidelined, effectively their community and projects were as well. They had to use the same software and attention to their needs has been inadequate.

Versions in other languages were started and each of them has its own history and community. All of them used images and all too often the same picture was uploaded. The cost involved was such that a new project started. Commons saved us a lot of money. Money that was initially provided by the founder of Wikipedia, Jimmy Wales.

Wikipedia grew spectacularly so much so that Jimmy could no longer afford it to foot the bill. The general public has been the main sponsor of Wikipedia ever since.

One of the best thing of Wikipedia are its links between articles and the links between articles in different languages. The links between the different languages were a mess and a new project, Wikidata, was envisioned to 'fix' this. Wikidata did so in a spectacular way; interwiki links are now manageable and this project is going its own sweet way.

All of those projects are works in progress. At 15 Wikipedia is very much an adolescent. A legion of loosely related projects and communities do their best to make it work. If there is one wish, it is that it will not be shackled by well meaning terrorists of reason. It was not 'reasonable' for Wikipedia to succeed and it is not reasonable that it has to fit in arbitrary molds.

Thursday, January 14, 2016

#Wikidata - Michael Milo, a Dutch #psychiatrist

Together with several others Michael Milo authored a book about mental health. Adding humans to Wikidata is trivially easy, Mr Milo is relevant as an author of the book so that explains the initial notability.

Given that mental health is not well represented on any Wikipedia, it helps when Wikidata has at least some information. Asking Mr Milo was easy; he can be found on social media, So the source of the information on his profile came in the form of personal communication.

A picture paints a thousand words and Mr Milo promised to upload a photo to Commons. Once we have similar information about all the authors of the book, it becomes easier to write about them on Wikipedia. All the authors are notable in their own right. It just takes someone to write or improve their articles.

Tuesday, January 12, 2016

#Wikipedia - #Recovery and #mental health

Recovery is not a method. According to Anthony it is a vision. Anthony describes the origin of recovery and it has not much to do with the 12 step program of Alcoholics Anonymous like the Wikipedia article says.

Even when you know all these things, starting to make changes is a daunting prospect. The article does not invite you learn more about recovery, it lacks focus and trivia like a possible etymology detracts from the subject. It becomes too long to read.

However, changing a Wikipedia article is frightening. Even when information is manifestly wrong, there is all the baggage required before you can edit an article. When I wanted to "fix" an incorrect link, I wanted to preserve the fact that the link was not a red link. I removed the #redirect and it was "wrong". I wrote a minimal stub, I even included a source and it was "wrong". The article is destined for deletion so the problem is likely to remain. Apparently nobody cares, I informed about the issue on talk pages; it took me already too much ..

I remember the days when it was welcomed to be brave.

The coverage of mental health is sub standard. My question is how to improve this. The question why it needs improvement is easy and obvious. The prevalence of mental health is such that high quality information is extremely relevant. When people are to recover, they have to rely largely on their own resources, their own ability, their own sense of self.

Working on Wikipedia articles is not my cup of tea. Working towards an editathon on the subject is. It will still have the same amount of stress but it will not be my stress. It will be real Wikipedians and people knowledgeable on the subject who will together make a difference.

Sunday, January 10, 2016

#Wikipedia - a 20% error rate

The one thing that makes Wikipedia strong are its wiki links. When they work they are great.. when they work they are.

The article on the Spearman Medal is a case in point. This medal is conferred by the British Psychological Society to psychologists. There were 19 links and two were wrong. One link was to a soccer and one to a football player. The award is conferred since 1965 so there ought to be quite a number of red links

With two sportsmen attributed to winning the Spearman Medal there was an error rate of 20%. With all the red links it is easy to be more informative using Wikidata. With such statistics it is obvious to make the argument that replacing links with links through Wikidata will enhance quality in the English Wikipedia.

This is unlikely to happen. Wikipedians seem to be more concerned in finding fault elsewhere than considering the quality of their own project. Particularly when "outsiders" point them to the error in their ways. It is psychology in action. 

Saturday, January 09, 2016

#Wikidata - William Anthony, PhD

Mr Anthony defined recovery and his definition is quoted on Wikipedia:
"Recovery is a deeply personal, unique process of changing one’s attitudes, values, feelings, goals, skills and/or roles. It is a way of living a satisfying, hopeful, and contributing life even with limitations caused by the illness. Recovery involves the development of new meaning and purpose in one’s life as one grows beyond the catastrophic effects of mental illness."
Recovery in psychiatry is proven effective. Mr Anthony does not have an article in any Wikipedia and, that can easily be understood as being part of the stigma that the subject holds.

In constructing a Wikidata item for Mr Anthony, there is the website on the Centre of Psychiatric Rehabilitation, it is part of Boston University. It has a page on Mr Anthony and it is definitely a source. Many publications and awards are listed and as you may expect many of those awards have no articles either.

In order to honour Mr Anthony, the awards had to be added. Several of the organisations that conferred the award had to be added as a consequence. That is where the page on Mr Anthony becomes a problem. A single reference to an award is not that great. But what to do if it is hard to find on the Internet?

Thursday, January 07, 2016

#Wikipedia - #Search - #Wikidata

Getting oodles of money to improve search is wonderful. A lot of good work will likely be done. It is also a great moment to consider existing questions that have not been resolved. It is all about having no result and how it might be improved giving the unique opportunities of the Wikimedia projects.

The first thing to consider is a favourite story of the first chairperson of the Dutch chapter. He volunteered at a library and people complained that they could not find what they were seeking for. He tracked what people could not find and realised that there is a people problem; they cannot spell. He aggregated the numbers and included a redirect for the really problematic words. People were happy. Librarians were happy because they now had a way of divining what books to buy on what topic; they knew what people were looking for and the library could not provide.

This is also possible for a Wikimedia search. When we know what people are looking for and cannot find, we can provide information anyway. The information on Mrs Boevink has been fleshed out but it only exists on Wikidata. When it has a Wikipedia article, it may not be in your language. It can already be provided as a search result but that means that "other" projects are taken seriously.

When a subject is not found at all, it means that people have an interest. It means that an article on such a subject will actually be read. This is a metric that people would enjoy; how well read are the articles that I started.  What impact do I have. It is only a thought but it will motivate people to write Wikipedia articles. This is particularly important to the small Wikipedias. When they write what people want to read, it will prove a process that positively affects the community.

So oodles of money have come in. Now lets be serious of finding Waldo.

Gerard de Reuver

When I was young, I visited my uncle for a few days during my holidays. He was a milkman in Haarlem, a gregarious person and he had his addresses where he went for a cup of coffee.

At one address, they had spirograph, a wonderful toy, I played with it for a few days and I loved it. When I went home, he "paid" me and instructed me to buy sprirograph for myself. I told him that I could not do this.

He called my parents and told me about it. They assured him that I could if I wanted to. I may have been nine or ten and, to me it was the kind of thing that is wonderful to play with when you come across it but not the kind of thing that I would always play with.

A good memory.

Tuesday, January 05, 2016

#Wikidata - Wolfgang Stroebe, not a Dutch psychologist

Mr Stroebe is an emeritus professor, a psychologist who worked for German, American and Dutch universities. From 1992 he worked at the university of Utrecht but that does not make him Dutch.

People could be mistaken for him being Dutch because the description on Wikidata says so: "Dutch social psychologist". Disambiguating Mr Stroebe with Reasonator shows clearly the qualitative difference between automated and manual descriptions.

The difference between the two is an old argument. People may like them but factually manual descriptions are inferior. It takes little effort to improve automated descriptions on a big scale while nobody really looks at descriptions of individual people like Mr Stroebe. Manual descriptions are translated and who dares to suggest that with translations the quality of the descriptions improves? The opposite is easy to argue. With translations, issues become even harder to remedy.

Never mind.

Monday, January 04, 2016

#Wikipedia - the lowest hanging fruit from #Wikidata

Contrary to what people think, at this time Wikidata could provide the most value particularly to the big Wikipedias by replacing the current red links. Red links are really simple. When an article exists, it will link it and that is exactly where things go wrong; when an article is added for a homonym, you need disambiguation. This is where Wikidata makes a difference.

Wikidata is ready because it includes all the articles of all Wikipedias and thereby its capacity to disambiguate is superior. Moving existing red links to Wikidata items may be automated because articles in other languages may have that link already. New red links in articles are checked for a need for disambiguation and when there is no need, a new Wikidata item is created.

In the Reasonator the concept cloud shows links as used in Wikipedia articles. Typically these links could exist as statements on the item. For Mr Ormel, a Dutch psychiatrist, there is only one article so it is not a complicated example. In the text there is (now) a red link to Mr Frank Verhulst. There is now an item for Mr Verhulst and as people convert red links, the statements on the item may be reflected in a tool like the concept cloud.

Another benefit is that links can be verified by tools like Kian. When a subject is about psychiatry for instance, a link to a person who was a football professional all his life is likely wrong and in need for disambiguation.

Arguably changing the red links in Wikipedia can be easily enhanced using information from Wikidata. It will improve quality in a meaningful and measurable way. Many of the functional parts already exist and it is therefore mainly a task of cobbling things together.