 Okay, thank you very much. Good afternoon everyone. Thank you especially to the organizing committee for accepting our presentation in this conference. We are very excited to be here. As we all probably know, open science tries to guarantee the transparency of research work sharing the different stages of the workflow. In recent years many organizations decided to support this kind of activities through the development of public policy or the encouraging development of public policy. But these and also financing open science initiatives. But these efforts vary greatly regionally and especially many times are not as engaging as the research community would like to. It has been observed as well that many communities have organized themselves to encourage this kind of activities and especially in some cases dedicated to reduce gender gap in steam areas or also to teach open practices and tools such as metadosencia does. In this context we wonder how these communities of practice influence the dissemination and implementation of open science in Latin America. To start we would like to introduce ourselves. Here we are Jessica Formoso, Mariela Reingeberg, Patricia Lotto couldn't make it personally but was part of our work and myself is Laurel Azenzi. We are part of accessibility, impact measurement and communication teams of metadosencia. Also Julian Buede who is part of communication team helped us with the slides. So let's take a closer look to this concept community of practice. Community of practice means a group of people that inays together to develop a collective process of learning about some different area of human endeavor. According to the Tiena Beverly Banger Trainers work it has to do with domain because they have a shared space or area of interest, competence and commitment that distinguishes them from others. They are also identified by building community through joint activities, discussions, problem solving, also sharing information and relationship building. A community of practice always encourages the idea sharing and fosters that kind of willingness to share knowledge and practice means that the members of a community are actual practitioners of this area of domain. They are going to come to the community, share ideas and knowledge and then go back to their practice and improve their performance. Also we want to address this dimension of today's communities of practice scenario which are the online platform channels that allows these communities and encourages these communities to expand their ideas and their work. As Schmidt and let me check on the surname Matilda's Akerland work show online platforms provide opportunities for participation and inclusion in new ways and are great to expand these communities of practice. Scholar communities have gathered for centuries, motivated by furthering human knowledge and development of humanity and these platforms sound great but these communities also struggle with the fact that the platforms are private corporate property and that creates new issues that we're going to address later in the conversation. Okay, to accomplish our purpose we conducted exploratory analysis of terms open science, science abierta in Spanish and science abierta in Portuguese through Google Trend Search and afterwards we developed a network analysis of tweets to see how influential users are connected to communities of practice. In this small graph we only wanted to show that in the context that some local researchers like Fernanda Begel and others are showing that the global consensus on open science that was encouraged by UNESCO's open science recommendation in November 2020 has reached to a region. There is still a lot of difference between the conceptual development of the term open science and the concepts ciencia abierta and ciencia abierta in Portuguese like the the line show. In the regional breakdown we can see that ciencia abierta is more used in Mexico, Colombia, Venezuela, Argentina and Chile and in Portuguese only relates to Brazilian territory. Thank you Laura sorry I changed your name. Okay so we conducted an analysis of tweets that were posted by a variety of users and that mentioned either as a hashtag or as a part of the text one of these terms open source open access open data open science and their traduction translations to Spanish and Portuguese bear with me with English and we use this to analyze if community communities of practice and their accounts had an influence on the conversation that was carried carried out. Okay so here in this graph you can see in this plot you can see how frequent each of these terms was how many tweets mentioned them and the more the most used ones were open source and its translation to Codigo abierto in Spanish. After one after sorry after we collected this information which was over 16,000 tweets in a period of a month we filtered this data to just retain those that were posted in Spanish or in Portuguese. We also just kept those that were related to personal accounts or communities accounts. We excluded commercial accounts and those related to government entities and we also tried to pinpoint the location of the user which was trickier than we thought initially. We used their location and the description of each user and we excluded those that explicitly mentioned that they lived or were originally from Spain or Portuguese and we assumed the rest was Latin American so it's a big assumption there and with that we had a sample of over 2000 tweets generated by 1400 different accounts and retreated over 7,000 times by 4,700 accounts approximately. We used this information to carry out a network analysis. Each user was a node and the links between the nodes the edges were the retreat so in this case we have a tweet posted by our lady Buenos Aires so I wasn't using the microphone back then right? Okay thanks. I hope you heard me and I retreated it. This is my account and Metalsensia retreated it also so we have three nodes from three different users and two links one between myself and our Lady Buenos Aires targeted links because I retreated something they posted and one from Metalsensia to our Lady's Buenos Aires. We have no link between each other because in that period I didn't retreat anything from Metalsensia. Metalsensia didn't retreat anything I posted with those specific terms. A concept we are going to use and I wanted to clarify briefly was the measure of between nest centrality because this concept shows us the number of times a node any node is in the middle of a path between two other nodes so it works as a bridge so in this case our Lady's would be a bridge between myself and Metalsensia. If I take this node our connection dies and that makes that specific node important to the creation of the network. So we took all those tweets and all those accounts and we classified them. We checked which accounts were part were accounts of communities of practice which was sometimes very easily identifiable and sometimes not so much and then we got metrics to try to compare how influential were the users from one group and the other. So we have average of retweets, percentage of accounts with original tweets that is the percentage of accounts in each group that has posted original tweets instead of just retweeting. We have average between the centrality which is the measure I mentioned previously and we have percentage of influential user accounts that I will explain how we extracted in just a second. So you can easily see that average of retweets is not that high but it's higher in community accounts and in personal accounts so maybe there is not that much movement with the terms that we search for but it's still more movement coming from the community accounts. The percentage of original tweets is a lot higher in community accounts and in personal accounts. Average between the centrality is also a lot higher so communities accounts are frequently in the path between two individual users and influential accounts. Well there was there's a lot of different criteria to measure how influential a user is and the one we chose is how many individual accounts retweeted that account. So that would mean that for example if I retweeted 30 posts that were posted by Metaocentia I only count as one point for that measure. We use as the threshold kind of 20 points and you can again see that the percentage of influential users, influential accounts, influential users in the community groups is a lot larger than in the personal accounts. We also wanted to see what happened with all the community accounts how they relate to each other and we took the ones we found in the previous search and we also added additional accounts that we know are from the region. This is, I know a lot of accounts, a lot of communities are not included. It was a sample and each link shows the number of followers they share and what we could see is that even though you can see there are some communities that are a lot that had links that are stronger that have more followers in common like OpenLavice and Charisma and others that have a lot less like El Gato y la Caja and I don't know our ladies or Datalat. What we found is that each community has at least one follower in common with each of the other communities with the exception of the one that's over there that is Connectorial. Connectorial is a very young community with very few followers and still shares followers with most of the communities. So it's, hi, okay, hello. So another analysis that we did was another network analysis where each of the nodes is represented either by a community or by a user account and the edges, we have an edge between two nodes if the user is following that community. Once that we did that network that we construct that network, we did a modularity analysis. This modularity analysis gave us some clusters. This means some groups of communities. For example, there are some clusters like this one that is over here that is more related to our ladies groups. For example, from Santiago, so Pablo, Río and Lima, we have other clusters. For example, this one that relates our ladies with metaosencia that is more related to geographic location from most of the people that form those communities. And we also have different clusters. Okay, thank you. Like, for example, Illa with Fundación Vía Libre and Carisma that are more related to ethics in data. That's what we thought about the relations between each cluster. So, as final remarks, we, recently I told you about modularity. There we have these different groups. Each group has like a dense, dense, like a dense connection, sorry, dense connection between the communities that are in that cluster and they have like, they are not so related with the other clusters or the other communities that are in each of the other clusters. So, these different groups, we find that we're mostly associated to geographic localizations and the purpose of each of the communities of practice. And then, after something before Shethi told you about the differences between the community accounts and the personal accounts and the difference in their retweets, each of one had and also in the proportion of influential users. And we found that in this exploratory analysis, communities have like a higher dissemination power of the terms related to open science. And then, with the analysis of the between the central measure, we saw that these communities of practice have a K-roll in the connections between users. That means that in generating this huge net. But we wanted to tell you some difficulties that we have in this study. The first and the most important was that when we start thinking how we could study these retweets was before Elon Musk had Twitter. It was to control of Twitter. And then what we happened is that we had a restrict access to the data. We couldn't have all the things that we wanted to study. This is something that happened with data that we wanted to do like time series analysis and those things and we had to restrict for a month. But we find some things that are interesting and encouraging. And other difficulties were like this one, perhaps in investigation something happens, sometimes happens, that is that, for example, the definition of influential user in the bibliography, there wasn't one definition. There were lots of definitions. So we used the one that we could measure. And another difficulty that also happens when we work with data, I think, is that we had to do like labeled by hand, which were related to communities and which were related to users. So this also was part of our analysis. We want to thank you all for being here. If you want to join us, you can go to metadocencia.org. We want to thank CSNS for being our fiscal sponsorship. And if you want to ask any question, we are here. Sorry about that. Thank you very much.