Wikidata power contributor: an interview with user Bodhisattwa

Background

Hi Bodhi! Thank you for your willingness to share your views and experiences on Wikidata and other projects. Can you give a brief overview of your efforts with Wikidata and other Wikimedia projects? 

I started editing in 2012, mostly on Bangla Wikipedia, where I focused on articles on Tibetan Buddhism. I was an administrator there for four years, after which I shifted my main focus to Bangla Wikisource, where I had helped a few other editors to build the project to what it is now, both as a regular editor and as an administrator. Apart from that, I have also spent time on Wikimedia Commons and Wikidata mostly around topics related to West Bengal;  my focus has been to use these two projects as a support platform for Bangla Wikisource.

You’ve still been very active in Wikimedia projects over the last decade. What drives you to continue contributing to Wikidata and other projects? 

Frankly, it has become a habit now. I have always loved contributing to my language since the beginning, and the love has not faded. When I started, content in the Bangla language was not widely available on the internet; that was one of the driving forces for me to contribute to Wikimedia. The internet has been on the forefront of knowledge preservation since its inception. In this world, if any of our languages are not present, they are bound to disappear sooner or later, no matter how hard we try to preserve them elsewhere. I was fortunate to find a bunch of passionate lovers of the Bangla language in my community from both India and Bangladesh who inspired and motivated me a lot. Now, when I look around, Bangla is way more prevalent on the internet than before, and I’m happy that I’m one of the many contributors who have helped to make it possible. But we still have a long way to go. 

I can only humbly request to give more love to our MediaWiki infrastructure to make it more user-friendly, interactive, and robust. We are lagging behind contemporary technologies in an alarming way, and if we do not become bold and invest for the future here, we ourselves may be history soon.

Wikidata user Bodhisattwa

Wikidata and other projects

Can you share an example of a project where you integrate Wikidata with other Wikimedia projects? 

These days I have been spending most of my volunteer time integrating Wikidata with Wikisource, particularly for the Bangla language. There are several inspiring volunteers from different communities who have been working on this integration for Wikisource.

In some parts of the world, where references are behind paywalls or not available on the internet, Wikisource can be a good source of references. You can find a range of reference materials, like several hundred volumes of encyclopaedias, dictionaries, chronicles, journals, newspapers, etc., which can be widely used as citations on different Wikipedia articles that lack them. Lovers of literature can find and enjoy a wide variety of fiction in different languages. 

Now, if we leverage the power of Wikidata for Wikisource, that helps make it smarter than before. We can always store the metadata of the Wikisource contents on Wikidata and roundtrip them back to Wikisource when and where needed, which can help fix errors easily without duplication of effort or redundancies. We can build a bunch of effective tools using the data to visualize the data in an interactive way to serve different purposes. Wikisource, being a project of unstructured texts, can also be structured in some way using Wikidata to make their actual text contents useful for machines. The potential is huge. Mahir Morshed and I have been working on different ideas to integrate Wikidata with Bangla Wikisource for quite some time now. These days we are working on a Wikidata-powered library catalogue built for Bangla Wikisource, which is being developed by Mahir over on Toolforge. 

Can you share some of the challenges you experienced when trying to synergize multiple Wikimedia projects, such as Wikisource with Wikidata? 

Wikisource has been a neglected Wikimedia project since its inception and has not received the love and care it deserves historically, including at its technical core. Not only does its interface look odd to book readers, but its infrastructure is also very old compared to contemporary digital libraries. In a practical sense, the volunteer communities need to develop everything from scratch, including the ways of integrating the project with Wikidata to harness its power. So if a language project does not have a sound technological volunteer, that is a major obstacle to any kind of innovation around the integration.

The inherent limitations of MediaWiki are also astounding. While it is easy to embed Wikidata query visualizations such as different kinds of charts, tree maps, interactive timelines, graphs, etc. to third-party websites, it’s not possible on our own wikis. Currently we are displaying the data in the most boring way possible. To overcome that problem, we are creating different Toolforge tools to serve our purpose outside of the wiki itself.

Do you have any suggestions for overcoming some of the challenges you mentioned above?

I can only humbly request to give more love to our MediaWiki infrastructure to make it more user-friendly, interactive, and robust. We are lagging behind contemporary technologies in an alarming way, and if we do not become bold and invest for the future here, we ourselves may be history soon.

You are arguably the ideal Linked Open Data Wikimedian. As someone who can integrate Wikidata, Wikisource, Wikimedia Commons, Wikipedia and even Wikibase, what are your thoughts on navigating the complexities across different projects? 

I am afraid there is not a single, simple answer for that. The complexities vary widely through different projects, sometimes among different language communities within the same project. For example, major Wikipedia projects typically have large- to medium-sized communities, and naturally it takes a huge amount of effort to discuss and gain consensus if you want to make a radical change, such as moving away from repeated and redundant manual data entry to a Wikidata-based entry system. Meanwhile smaller language Wikipedia projects try to adapt Wikidata into their workflow to avoid these repetitive tasks to preserve their precious manpower, but unfortunately if they lack the necessary technical expertise, beyond a point it becomes difficult to experiment. 

On the other hand, as far as projects long neglected from the top, almost all Wikisource language communities were on their own from the start and eventually became highly flexible regarding experimentation and changes, with a culture of sharing learnings among each other across languages. Whenever they managed to gather a few technical experts, they could also convert their ideas into working tools, many of which are directly related to integration with Wikidata on many fronts beyond the usual Wikidata infoboxes and Listeria lists. 

Somehow Wikimedia Commons stands in between. The community is flexible enough, implementing Wikidata infoboxes early on and allowing different campaigns to run on Wikidata-based lists. When SDC arrived, it came with very high expectations regarding potential ways of integration with Wikidata, but the authentication feature of WCQS has severely restricted its full potential and discouraged users to even think about SDC.

Regarding Wikibase, I have only spent a few months playing around Wikibase Cloud and not the Suite. I think Wikibase Cloud requires more robust and user-friendly features like batch uploading of data and linking with Wikidata. The tools used there glitch often, and some power tools cannot be used right now.

Final thoughts

Finally, what advice can you give to the newbies of Wikidata, Wikisource and Commons?

If I had to convince someone to contribute to one of the Wikimedia projects, I would suggest always keeping in mind the quality of the content and ways to improve it. It’s a difficult task for newcomers, who tend to focus mostly on quantity, but it is always sufficient to add verifiable neutral references on Wikipedia, proofread meticulously on Wikisource, add statements following established data models on Wikidata, etc. — whatever seems convenient for the users. These quality edits cannot be easily counted on our statistical tools for contributions, which newbies like to check a lot, but I would suggest simply overlooking them.

Thank you so much, Bodhi, for everything!