#Wikidata items without a statement

Most problematic in Wikidata are the items without a statement. They may be connected to articles in many projects but the main issue is that it is hard to add statements with confidence when you do not know what they are about.,

I blogged about datathons and one result is a new tool by Magnus that can be of an enormous help; It shows items without statements for a language and they are sorted by the number of sitelinks they hold. There is a link to the Reasonator and it is really useful because it shows some text when you hover over a sitelink.

It is yet another tool that is worthy of attention in Mangus's tour of the Wikidata ecosystem. The example you see above is for Dutch, this one is for Nepali.
Thanks,
       GerardM

#Wikidata - What to do in a #Datathon III

One goal for a datathon is to demonstrate, to teach the use of the various tools. When you want to be effective individually or as a group, you have to use tools. The Wikidata statistics make it obvious; the numbers of items, statements and labels are huge. To have an impact it is important to use the tools that are available.

The mathematics involved explain how you can have the most impact. Set theory helps you understand how to approach this effectively. When you add one label and it affects 10,000 items it is powerful. When you add every known lawyer of a country in one go and it affects 10,000 items it is powerful.

Choosing the tools that help you be effective is therefore what you teach the participants of a datathon. They go home and they are likely to continue using the tools.

The most obvious tool to introduce is Reasonator. For it to function well, it needs to be configured. This is done partly from the personal settings. This ensures that you Wikidata data will show as information in Reasonator in *your* language by preference.

The other part to configure Reasonator is Widar. Widar is a tool that authenticates the edits you do in Reasonator to Wikidata. In the screenshot above there is a picture that was added to Wikidata in this way. The most relevant use case is adding labels in your language. The missing ones are underlined in red.

AutoList is probably the second most relevant tool to demonstrate. It provides an interface to several tools. One of them is the WDQ or Wikidata Query another is the CatScan. Combining these tools allows you to find every human in a category that is not known in Wikidata for what the category indicates is true about this human... and then add the statement that does just that.

There are other tools that are extremely useful. In a datathon it is important to have a mix of people. People who know the tools well, people who are interested to learn and people who have a "mission". When the activities concentrate on the "missions", it means that you zoom in on a subject and apply the skills and labour on doing good. This makes it clear how relatively little effort can have a huge impact. Not only for a language but also for a subject domain.
Thanks,
       GerardM

#Wikidata - What to do in a #Datathon II

There is little point to a Datathon when the results have no practical impact. By implication there is little point to Wikidata when it has no practical application. Luckily most of the Indian languages use Magnus's extension to search making any and all advances in Wikidata immediately useful.

The next thing is to decide for a datathon is what it is you want to expose in your language, your script. The result will be biased but the difference is in not sharing this information. That option is even worse.

Having said that, there is one upside to concentrating on a subject domain. Take for instance the "King of Nepal", you will see that it is referred to "List of monarchs of Nepal". All these listed monarchs are now a "king of Nepal". It now takes one person to add a label in Nepalese to make this label visible on all the monarchs of Nepal. It is a subclass of "king" and, it takes one person to add a label in Nepalese for all subclasses of king.

This is the beauty of adding labels in Wikidata. Once it has been added, it is used everywhere. A label for "politician", "lawyer", "date of death" are added once and are in use on hundreds of thousands of items. Adding labels is therefore really satisfactory and effective.
Thanks,
      GerardM

#Wikidata - What to do in a #Datathon

I was asked for pointers for a "datathon". It is adding data to Wikidata for a specific purpose. The most obvious thing is to be clear what it is you want to achieve.

What to add to Wikidata:

• adding labels to items in a language
• adding statements for existing items
• adding items and statements based on a Wiki project
• adding missing items to create links among items

Realistically, it is always a bit of all of that. The people attending are not all the same either, they differ in interest and they differ in skills. One goal for a "datathon" may be the transfer of skills. When this is the case, start with the basics of Wikidata. How to add labels, how to add statements. how to add items. 

Another goal is to add information for a specific domain. This may be based on information known to a Wiki project but that is optional. When information for a specific domain is to be worked on, Working together and use as many tools as available makes a real difference.

As a blogpost should not be too long, more later..

Thanks,

      GerardM

Condensing the concept cloud

The concept cloud is a #Wikidata tool that indicates what #Wikipedia articles link to a given Wikipedia article in any and all languages. The idea is that is will help you decide what might be included in a language.

It also allows you to check the labels in your language. Make sure that the spelling is correct and add any missing labels. Quite often you find labels that have not been linked yet. In the screen shot you find items in Nepalese.

Linking them up is part of creating an improved concept cloud. What is really gratifying is when you find halfway that other people are working on merging items as well.
Thanks,
      GerardM

The #Bihari #Wikipedia is actually written in #Bhojpuri

This is the kind of article that has many people's eyes glaze over. It is about standards, scientific documents and it is about languages most of my readers have never heard about. For the people that do speak one of the languages that are considered Bihari it is extremely relevant and it has implications for Wikipedia.

This is information provided by Umesh Mandal that explains about the "Bihari group of languages" in relation to the Maithili language:

Kellogg (1876/1893) and Hoernle (1880) regarded Maithili as a dialect of Eastern Hindi; Beames (1872/reprint 1966: 84-85), regarded Maithili as a dialect of Bengali, Grierson has done a great service to Maithili language, however, he erred when he gave a false notional term of "Bihari" language, after that western linguists started categorizing Maithili as a dialect of "Bihari" language; although there is nothing known as "Bihari Language" and both Maithili and Bhojpuri are spoken in Bihar (of India) as well as in Nepal.

Umesh is working on the localisation of MediaWiki for the Maithili language and as this language is currently in the Incubator, the language committee does its due diligence and tries to understand if Maithili can have a place in the Bihari Wikipedia. The information provided by Umesh makes it quite clear: "no".

This still leaves us with the misnomer that is the Bihari Wikipedia. Apparently the language used for the localisation and the articles is Bhojpuri. Bhojpuri has the ISO-639-3 code "bho". 

Are you still following all this? Ok, there is one question I am not asking: How about the Kaithi script?
Thanks,
      GerardM