International Wikidata

Editing Lexemes with your little finger

21. March 2025
Dan Shick
The Lexica logo
The Lexica logo (Kenny (WSC), CC BY-SA 4.0, via Wikimedia Commons)
International Wikidata

We’re always looking out for ways to empower users to work with Wikidata more easily, and in today’s interview we got details on one of the most exciting tools we’ve seen in a long time. We were lucky enough to speak with Raisha Abdillah, who leads the Wikidata Software Collaboration Team in Indonesia; she shared some valuable insights about Wikidata informed by the project they’ve been working on, Lexica. 


Wikimedia Deutschland: Please introduce yourself and tell me a little about your project group.

Raisha Abdillah: My name is Raisha, and I’m the project lead for the Wikidata Software Collaboration team in Indonesia and our team. Lexica is part of the software collaboration project with Wikimedia Deutschland, and our aim is to create a tool that can empower the Wikidata community, especially the community here in Indonesia. 

Our team consists of a project lead (me) as well as one product designer, two community communication staff, one front-end engineer, one backend engineer and one person handling the administrative aspects of the project. That’s all of us. [Ed.: The full team list appears at the end of this post.]

As the project lead I coordinate everything and also handle the tasks that no-one else wants to do. *laughs*

WMDE: Seems like you’re doing a great job. So what’s the background of your work with Wikimedia? What did your relationship with Wikimedia look like before this project?

Raisha: I was actually a Wikidata volunteer. I started editing Wikidata in 2016 as a student, and I was working with the local chapter in Indonesia as a technical staffer since 2018. In that role I held workshops about Wikidata and shared my expertise with other community members about how to use tools and how to add data to Wikidata itself. I also managed several technical partnerships back then.

Then, in 2022, when the Software Collaboration for Wikidata Project began, I became product manager for the software team itself, which I built from scratch. I hired people and tried to get people into the project. We ended up separating from our local chapter, but as a separate group we’ve been working closely with WMDE to build Lexica.

WMDE:  Where did the idea for Lexica come from? 

Raisha: We did research in the Wiki community in Indonesia, three different groups: newcomers, casual editors and long-timers. What we learned is that there’s a need for a tool that is mobile-friendly, that can be accessed everywhere and anytime, and also a tool that can help people not to be afraid of Wikidata itself. There is a stigma around Wikidata — that it’s highly technical, that someone might break things if they edit it, or that you need to be able to code to edit it. 

We didn’t make Lexica with gamification. This might seem counterintuitive; sometimes people want to contribute hundreds of things. What we try to encourage is that people take their time, steadily and slowly.

Raisha Abdillah

This assumption very much also exists in lexicographical data. Even people who know they can edit films, books or other things that they can relate to, when it comes to lexicographical data they think, “Okay, I’m not a linguist, I shouldn’t touch this aspect of Wikidata.” So we tried to find ways to help people not intimately familiar with Wikidata make edits that conform to the structures already defined by the data’s maintainers. We wanted to make sure that the newcomers didn’t just edit things, but that they can edit the right things in lexicographical data. 

This is really hard to do because of that stigma. We needed to design things so the edits stayed within certain guardrails. That was the pain point we tried to address. 

The country of Indonesia has over 700 languages, and some of them are becoming extinct. When we talked with the community, we learned that they had a lot of trouble finding resources, even a dictionary. We saw that lexicographical data could be really useful here.

WMDE: How does Lexica work? Feel free to go into detail.

Raisha: I’d like to start with the vision itself behind Lexica and give a shout out to Kenny, our product designer. He’s been shaping how we see tools and how we build them in the first place. 

I myself come from a maintainer background, so I had the idea that a tool should have a lot of functionality, and it should feel like it does. However, from Kenny’s design perspective, a tool should have the power to guide people to do something right, do it mindfully. That stuck with me. I used to think that could only be achieved by making sure the tool has a lot of buttons. Kenny said no: simplicity is the key… but creating something simple is actually very difficult. So that’s where it started.

Screenshots from the Lexica app
Screenshots of the Lexica app
(Kenny (WSC), CC BY-SA 4.0, via Wikimedia Commons)

On the development side, we observed that within the Wikimedia ecosystem there’s a certain blind spot — that many actively-used tools cater to maintainers, who already have a perspective on what they’re working on, but that there aren’t many tools that cater to the needs of newcomers. And the tools that do exist are too simple; they don’t conform to the actual data model created by the maintainers. There’s a constant clash between these two kinds of tool. So we tried to balance these two perspectives, for maintainers and newcomers alike. 

When we created Lexica, we wanted to guide users in making a thoughtful and focused contribution. One of our principles is deliberate limitation. 

For example, for English Lexemes, we have just one contribution type: “link Lexeme to item”. We want to make sure that user feels focused when doing something. They feel a guardrail in what they are doing. 

Each card shows Lexemes not yet connected to an item on Wikidata. Why these? Sometimes users have difficulty deciding what to do next, so we give them something random to work on that’s in their language. We also give them recommendations, based on searching for the items with the same label. They can also search directly in Lexica. 

Part of the workflow for adding to “blossom” in Lexica (Kenny (WSC), CC BY-SA 4.0, via Wikimedia Commons)

After five cards, the session ends, and the user can go back to the home page. We didn’t offer a lot of cards in the beginning or create an infinity loop; you want people to just contribute small amounts of things. We didn’t make Lexica with gamification. This might seem counterintuitive; sometimes people want to contribute a lot, hundreds of things. However, what we try to encourage is that people take their time, steadily and slowly. Then they can start again after five.

WMDE: What kind of feedback have you received about Lexica? 

Raisha: We’ve received feedback from newcomers who are directly using Lexica. They’re using Lexica consistently. One user from Indonesia who had never previously contributed lexicographical data is now creating new connections daily with Lexica — we’re very grateful they’re doing that. And we also received feedback from maintainers editing with Lexica, and what we learned from them is that each language has its own way of formulating ideas about certain things. For example, in German, we had to start filtering out recommendations, because there are nouns that shouldn’t be directly linked to an item. German has a specific word for, I think, “midsummer lunch”, something like that, and that would not have a corresponding item.

WMDE: Because of compound words?

Raisha: Yeah, because of composita! So we talked with Wikidata maintainers about their concerns and about their views of their own language. It’s been great for us. 

WMDE: Do you use the Action API or the new REST API?

Raisha: The Action API.

WMDE: Do you have plans to use the REST API? 

Raisha: Not yet. Linking Lexemes to items, we’re waiting for that. We want to improve the search and the recommendations itself. The Action API still does its job, so we’re focusing on the search API first, and then we’ll see whether the REST API is more efficient to do what we currently do with the Action API. 

WMDE:  What else do you have planned for the future of Lexica? How would you like to see Lexica transform how people see Wikidata?

Raisha: We plan to add another type of microcontribution: the hyphenation aspect that can be added into Lexemes. It’s very easy — users don’t need to have an advanced linguistic background to do hyphenation. 

We also plan to add accessibility improvements for Lexica, reducing motions, things like that. We want to support the internationalization aspect of Lexica, like RTL languages… also maybe putting Lexica on translatewiki, so other people can translate the interface as well. 

And then maintaining documentation. We want to make sure that Lexica is stable. Even though it’s doing small things, it should do small things right.

We really hope that Lexica can inspire the way people work with Wikidata, so that they’re not afraid of doing it, and also inspire how the way people develop tools for Wikimedia. We want people to be thoughtful in creating things and also be mindful of other contributors. We want to minimize the knowledge gaps that we currently have in Wikidata. The way to do that is by making sure that the tools are friendly enough to contribute. 


If you’re interested in learning more about Lexica, you can read all about the project or dive in and try out the mobile-friendly interface.

Thanks to Raisha and the entire Lexica team!

NameRole
Raisha AbdillahProject Lead
Kenny TjahjadiProduct Designer
Harri RahdianProduct Owner
Faridh MaulanaBack End Engineer
Hendry VariantoFront End Engineer
Kartika SariCommunity Communication Staff
Ivana LiviaCommunity Communication Staff (past)

No comments yet

Leave a comment