 recording going and the recording is on and I think a way to use. So I'll hand it over guys. Thank you. Yeah, thanks. Thanks a lot and thanks. Thanks a lot for organizing this event and keep the controls so tightly in your hands. It's a really excellent job. Therefore, the organizers are amazing, amazing. My name is Gabor Kishmihova. I'm the head of learning and skill analytics research group, which is the LiveNet Information Center for and yeah, welcome to my home office. We have been stranded here for about 20 days. And according to the government, we have 80 more days to go. So I apologize in advance that if you hear screaming around me, those are my kids, but I guess this is normal in these circumstances. So today, I want or we want to talk about OER recommendations to support career development of individuals. Projects, part of a European Union, European Commission finance project, but it's a bit bigger than that. Institutions to work on this project. One is TIB. We lead the project and the other one is University of Amsterdam. Stéphane Moll and Ellen Berg are the ones who participate in that team. And our point of departure. I don't think I have to introduce realisation here in this group and how educational technology personalised education and learning. We believe that learning is getting more and more personal. It's driven by a great number of individual goals. We have the opportunity to sort of control and react to those individual needs. The big question, what I've been facing in the past years, that actually who's responsible for learning when it comes to personalisation? So mostly we are active in the higher education sector. So what we do is mostly focus on higher education and especially in the Netherlands and in Germany. Responsibility is mostly on the institute of teachers. There is very little lies on the shoulders, which if you think about it, it's very strange because once these people leave academia, once these people leave higher education and they start work somewhere in these two countries or wherever around the globe, suddenly they become responsible themselves for their own career development, for their own self-development. And this is not really learning, actually at least not in our settings. So we want to problem and we sort of decided in this project to try to empower learners a bit and empower learners through open personalised learning and curriculum recommendations, which reflect on labour market needs and which vast amount of open educational content across the globe. So more specifically with, and this is a research project, so we have a number of scientific objectives, what we need to keep in mind. One is that we want to contribute to the understanding of learner behaviour through goal setting. So we want to look at the efficacy of goal setting in education. Surprisingly, there is very little amount of research happening in this area. Then we also want to contribute to the literature of self-regulated learning. We're developing novelised curricula, understanding the difference between individual, between the individual pathway of different individual learners, and looking at the attitude and changes in learners when it comes to this type of technology. And last but not at least, we also want to introduce methods which pull in non-educational data sources to education, especially in learning content development curriculum design or learning evaluation. We also have some very practical objectives that's basically building a recommender system, which suggests relevant open learning content to our learners. We want to make learners to target skills and jobs they like individually, and they want to master individually. And we are learners during the projects by providing visualisations and individual feedback. So how do we want to do this? Our concept image, and basically we want to start the process with goal setting. So all learners who interact with our system should set a goal, a skill or a job they want to master or they want to do after their studies. This job, we try to components of that job. How do we do that? We text mine vacancy announcement, official occupational classification systems, and the two sources of information. We sort of distill all those skills needs, what the students need to master in order to be the students selected. Then based on this information and based on the profile of the person, so what educational background this person has, where is the physical location and so on and so forth, we create a learning pathway for this person and we feel this learning pathway with educational content. This is coming from educational resources, but we also think about the content for this purpose. And last but not at least, we use analytics to provide feedback his or her progress and learning pathway. This is very important because goal setting is not a static activity. It's not only once, but as you progress towards your goals, you need the opportunity to reflect on your goals or even change it completely. Maybe after one year or so, you decide, okay, this is not my dream job and when writing a new learning pathway, we can consider all what you learned for your new targets, for your new objectives. So this is the general concept and now I hand over to microphone to Reza, who will tell you the technicalities and show you the prototype to make this idea come true. So Reza, the microphone is yours. Thanks Gabor. Hello everyone. Yeah, based on the concept Gabor just explained, we have built our first prototype focused on data science related jobs, according to the huge changes in skills demanded by labor market. We believe that OERs have potential to handle the increasing need for education, since they are provided by people in different contexts, such as country, city, job, and so on. So we have implemented our approach based on OERs. In our dashboard, users can select their target job, which can be their current or future job. Afterwards, we show them the required skills for that job. To match jobs and the required skills, we analyze job vacancies, as Gabor said, using text mining approaches. After that, we ask users to set their expertise level for each of the skills. So according to the expertise levels, we built a learning profile for users using similar learners, such as the ones with same job, same country, same city, and provide personalized OER recommendation for each of the target skills. After finishing each of the recommended OERs, we ask learners for their satisfaction and update their preference profile based on their rating. We also update users' expertise level as well until they get mastery level in all of the target skills. Here, we have some screenshots from our dashboard. As you can see here, users can search to find their target job, set their expertise level in each of the required skills that we suggest to learn. Here, you can see the learning dashboard where users can find the recommended OERs and their details. Also, users can change the recommended OER whenever they want. Moreover, this page shows the users' goals, reports on the finished OERs and their progress. For instance, the user reached his or her goal in data visualization, as you can see. We validated our first prototype focused on data science related jobs by semi-structured interviews with 15 PhD students and eight university instructors in the area of data science. In the first 15 minutes, we presented our research problem and the proposed approach. Afterwards, we asked them to use our prototype at least for 15 minutes and finally, we interviewed them using a questionnaire which asked about our objectives, our logic, and our contribution to learning. Now, I'm going to show you the output. Regarding the objectives, the interviewer mentioned that there is a potential value in building a labor market information based on OERs. Also, they told that finding useful OERs are complicated and time-consuming, so high-quality recommendation and services are required. They also suggested that we should also cover people who want to learn specific skills without targeting any job. For logic, they told that with this approach, we help learners focus on the most important elements of their current and future jobs. Also, our process for extracting learner's properties, grab their attention, and they suggest us to improve our assessment parts by asking technical or non-technical questions. And finally, regarding our contribution to learning, the interviewer mentioned that interacting with learners in order to recognize their preferences is one of the most important and engaging components of our proposed approach. Furthermore, they taught that setting specific and personalized goals for each skill in our prototype system has a strong positive effect on the learning process. And they told that we should capture more learners preferences, such as their favorite types of OERs, for example videos, presentations, pictures, and so on. Therefore, as the next steps, we are going to add more OER repositories, extracting and extracting learners' properties. Also, we should decompose skills into meaningful components in order to have better assessment for learners' levels of expertise and build more suitable learning paths. And now, I give the floor to Gabor again to conclude our presentation. Sir, I just want to wrap this up very quickly. We already have some publications and resources available. We put them here on this slide. So later, you can check this paper out if you want more information and want to go a bit deeper in this topic. But of course, you shouldn't hesitate to contact us through email or through Twitter. As we will make this system public in the upcoming weeks under OER-recommander.com. So it's going to be a public service, what you will be able to use as well. As as I said, our first deployment will be in the area of data science, but we are planning to open it up to other areas as well quite soon. And last, at least, thank you very much for your attention. If you can and if you're interested in this topic, please provide us a feedback. We have a short survey where we usually use to capture the feedback of audience and other relevant stakeholders. So please don't hesitate to take this link or capture this QR code and give us some feedback about this topic. So thank you very much for your attention. And if you have any questions, please let us know. We are happy to answer them. Wow, thank you very much for a wonderful session. That's really, really interesting. And we will, obviously, everybody will be able to see the link in the chat as well.