 actually entertainment session and would like in this session actually to highlight some of the aspects of the newly actually developed open Wikimedia Research and Development Ecosystem in Africa, my continent, as I mean from Tunisia. So this project actually is a part of the adaptability data to support clinical practice using data science, cement web and machine learning project. All right. So how this will work? Actually, there is a quiz of eight questions. We'll see how the audience will guess answers. I actually will provide the explain the answers. So that's all. Yeah. So actually, question one, how many African countries are involved in Wikimedia Research and Development according to you? Then then you can who says 10 raise his hand? No? 20? Yeah, the answer on the side. 20, 10, 10. Yeah. The answer is 30, actually. Yeah. Yeah. So actually, actually countries from Eastern, the Eastern coast of Africa are actually very involved in research. Actually, they start from Egypt to South Africa. All that goes actually is a high involved in research. And that's because there are some policies in Kenya, in Rwanda, in that region, actually, in Egypt and in South Africa that encourage AI research, and as Wikimedia resources, Wikipedia, Wikidata are free resources. So they are using that. And that's why there is an increase in that region. And there are the classical countries that are doing research in computer science. So there are Tunisia, Algeria and Morocco and Nigeria, they are classical countries that do research kind of for for decades now. There's also Ghana, Bali, etc. that are actually doing some computer research thanks to collaborations with the UK with France, etc. And on the other side, actually, there is a community called Africa Wikimedia Technical Community that includes actually communities from Tunisia, from actually Kenya, and from actually the African coast, actually Ghana, Nigeria, kind of. So all these countries, this is mainly linked to the work that has been done by the Open Knowledge Foundation, West Africa. So it is a foundation that has been created a few years ago, mainly by Felix Narty. And it involves Wikimedians that are interacting with the developers community in these countries in Western Africa, so that they can be involved in media, wiki development in wiki data, mainly script development for Wikipedia, etc. So the question number two, as we are talking about the creation of the African Wikimedia Technical Community in Ghana, when this has been created? That's question number two. There is 2016, 2017 and 2018. 2016. Who says 2017? 17. Yeah, I'd say 17. Yeah, and actually, as I said, that that is the work of Felix Narty and Bersy Zali. And actually, they put strap this at the Wikindaba conference held in Tunisia in 2017. And actually, the kind of establish good relations with open science communities, and with with with with with computer development clubs, like Google developers clubs at universities. So they are doing doing a very excellent job. Yeah, so the question number three, actually, as there is a community, can you name five active African developers working on a community project? And most of us, most of us will say no. So that these are actually five active developers that actually some of them are in Wikimedia this year. So you can meet my head on my tongue is from Benin. And he was from that community. But after that, he was recruited to work at the Wikimedia Foundation. And now he's among the staff of the Wikimedia Foundation. He's doing some really excellent work regarding media wiki, mainly development. API is all the staff related to media wiki. And there is also an electric was the first to participate at the Google summer of code. So there are some other people, like Iqbal, Eugene Akbar from Cameroon, some oil, Nigeria and Abel Bula, that are mainly working on Wikipedia script development. And some and now they are shifting to wiki data related development. Yeah. So where is the most productive research institution working on Wikimedia projects? That's for sure. Yeah. So actually, our research unit is the one doing that. And it is located at the University of Sparks, Indonesia. This is our headquarters actually, as you see on the right. Well, we were formerly working since 2011, actually. But we are we were recognized in 2021, as a nationwide research unit that is not 2012. Actually, we have now, thanks to Wikimedia Funding and Wikimedia Research Fund, we have the headquarters and we have the data center that can work, distributed computing, some fancy machine learning stuff, and some DevOps and DevOps. So we are kind of good in things like that. And we have 29 research publications around Wikimedia research as per scopes. So if you see YouTube, we have a presentation at the wiki workshop of this year about what we do, how we work on that. And there's also the URL that you can check. There's also our business card where you can find the same, the same QR code involved with the URL. Yeah. So actually, question number five. Do African countries integrate the forefront of Wikimedia Research and Development? And I think that's a pretty simple question, actually. The answer actually is no. They are behind the scenes, actually, they are mainly dependent on collaboration. So as you can see, there are quite recent efforts for the integration of the African countries, and they are mainly dependent on international collaboration in set it, particularly like the ones with Wikimedia Deutschland, et cetera. So mainly occurred thanks to collaboration with the BRICS country, which is quite surprising. So the countries that are mostly contributing to Wikimedia research, like Germany, and like the United States of America, were not the people that pulled out African countries to work on Wikimedia Research. These were BRICS actually countries like Russia, Russian Federation, Brazil, India, China, and South Africa, because it is in Africa. So that's quite evident. And we have quite a poor productivity of African countries. Unfortunately, that's the situation. So if you see here, actually, the blue nodes here in the map are the old one. And the yellow is the new one. So if you see here, this is the Wikidata Research Collaboration Network. If you see here that Germany, Italy, the United States actually joined later. So all these have traditions in Wikimedia Research. But when it comes to BRICS, they are quite recent. They came first, and then they grabbed the African countries later to work with them on Wikimedia related research. Yeah. So what's the main target? Wikimedia project for research and development efforts. It is comments, Wikidata, Wikipedia, and Wikimedia. Well, the answer is Wikipedia. Yeah, that our work as a research unit, actually, among the 29 papers on Wikimedia Research, we actually have around 13 or 14 related to the Wikipedia category graph alone. So around nine now, we have nine published paper around Wikidata only. Around all of Wikidata with all its features. So that's why I think. Yeah. So why is Wikipedia actually? Actually, Wikipedia have plenty of open source codes online. If you go to GitHub, for example, you have 2.9 thousand GitHub repositories, open access on GitHub, reusing Wikipedia. This includes libraries, kind of scripts, etc. And we have 12.1 thousand scholar publications, according to scopes. Particularly, they're the Google tech giants, the tech giants like Google, like IBM, develop Wikipedia tools that are not involved in Wikidata development as it should be. There is a huge interest from research institutions all over the world. And there is a small involvement of Wikimedia volunteers. And that's why if the community in Africa is there to establish some something, it will not help. And that's why they go to bricks, actually, to get some help to bootstrap Wikimedia research at that instant. And that explains the part how Wikimedia community as a community are separate from the research community. Because we don't have the people who have the double helmet as Wikipedian researchers. However, concerning Wikidata, there are 36.2 GitHub repositories, 945 publications. And it seems that it allows better engagement of Wikimedia volunteers and affiliates research and development related to Wikidata. And that's why we find a lot of Wikimedians that do some successful research works in scholar publications related to Wikidata and not related to Wikipedia. So it is growing. But actually, we need actually, if you want the Wikidata research to grow in Africa, we need some collection between the community that masters in some way Wikidata. And the research community and government within the triple helix collaboration framework. So well, question number seven, what is the most efficient governance of nonprofit Wikimedia research and development communities in Africa? So there are three models, actually, for, for, for no, for the kind of management of, of nonprofit research communities in Africa. There are the top down governance at all of our working with Wikimedia researchers, by the way. But there are the top down governance scientific societies that have shoppers, working groups, etc. We have the self organized distributed governance, what is called the grassroots. And there is the affiliate governance, which is Wikimedia affiliates that has a subgroup related to research, to Wikimedia research. So which one is the more effective in Africa? The most efficient governance, number of papers issued from these collaborations, number of grants that have been issued by the Wikimedia Foundation, etc. So the answer is indeed grassroots. And if you see, for example, Masakana here, Masakana is the African grassroots community for NFP research, and it had been awarded the Wikimedia Foundation research award of the year in 2021. And they have been funded this year, actually, for a project at the Wikimedia Research Fund. So kind of love them, if you like. Yeah, the second one is Cisoncabutec, which is the African grassroots community for medical machine learning research. And we have been establishing as data engineering and semantics research unit collaboration with them. And we have been funded the project I have advocated in at the beginning of this talk, with quite $50,000 as well. So last question. Well, is the African artificial intelligence community interested in Wikimedia research and development? In general, we talk about the computer science community in general, not the Wikimedia one. How they perceive actually Wikimedia research and development? And the answer here is, of course, no. Yeah, we have done a bibliometric study actually, around 90 pages, about biomedical machine learning to see what are the main topics that are advocated by African researchers related to this topic. And we have seen that most of them do not interest in open resources at all, not only Wikimedia projects. And we have seen that they are mainly interested in working on raw data. I do some few short learning some kind of trendy things, like, like neural network, learning, like decision trees, like, like some internet of things, think, think for patient monitoring, etc. So all things are mainly about about actually processing raw data, and not about doing things using on semantic analysis. Which is actually the good thing about using Wikimedia projects. Actually, why is that happening actually? Because using pagemark, rather than open resources, prevents the users actually to go through preprocessing. As you know, Wikidata or Wikipedia are our community generated resources at it. They may actually include some mismatches, and they don't like to lose time, actually. And because this is technically challenging, and because actually, as you know, that the training of PhD students is not that robust enough, actually, to go through this side, actually work to achieve the goal of their PhD journey. So that's that's the main thing. Yeah, so I hope the session was actually kind of light. I did not take much of your time. So this is my contacts. You can actually find my business card of myself at the head of our research structure at the back on the left. So you can search them as well. Thank you. Has anyone any question actually? Yeah, go on. Yeah, thank you for the very interesting presentation. Yeah, it was new to know that so much is going on in Africa. So really thank you for that. My question would be you are you were focusing on AI? Or is it so you can exclude it all other research which could not directly being related to AI? That's a good question. Actually, should I ask the second question? The second question would be, where did you publish your applications? And did you have some quality measures? Or how did you master that? Thank you. Yeah, the first question actually 90% of the research around around media in Africa was actually related to computer science. That's that thing. But there are there is Peter Gallert, I have seen him today from the he's working in Namibia as well. So actually, he did some work. He did some work about the social perspective of the contribution to do media research. But that that's one person. He published actually around 10 or 11 works on this purpose. Because certainly the second question actually, we published in highly referred journals. So myself actually, I have a paper published in Journal of Medical Informatics. I have a paper in Somatic Web Journal. I have paper in PhD computer science. Yeah, they are peer reviewed. They are high because I'm not a fan of impact factor, but they they are supposed to have a very high impact factor. So so they are good papers. Any question? Yeah, that's good. Yeah, so if you don't have any further question, yeah. Thank you. How do you explain that African research are not interested by the media project? Yeah, that's because first, the the their topic of interest actually, their own languages are not well represented in Wikipedia. And if they have to to actually do some NLP stuff with, for example, Europe by Wikipedia or something, they need a lot of preprocessing. And actually, there are some benchmarks that are quite easy to process. So actually, they go for the easy path. And that's all. So that that's why actually, we don't have that thing. Yeah, at the previous slide, previous slide. Yeah, I quite understood very I'd like to know it better. You said that the raw material, they have so they have to delete noises, and some type of works they require. Yeah, we could data actually is structured. And I think it could be all material for the training. But you said that I couldn't. Yeah, yeah. So the work that is done on machine learning in Africa is mainly is mainly considering the noise inside the input data. So they are, they are putting all the things inside the input data sets, and they are processing it. And that's all. So, so raw stuff with the very, very huge machine learning texture. And that's all. Yeah. Thank you very much. So I will leave it for the game.