 Hello, everyone. My name is Wei Junyuan. I'm a PhD student at the University of California Irvine. I'm presenting on behalf of my wonderful team, I'm consisting of Zhao Zhili at Washington University in St. Louis, Julia Wagner at ETH Zurich, Seth Frey at the University of California Davis, and Benjamin Macohill at the University of Washington Seattle. The title of our research project is One Path or Many, Policy Development and Diffusion Across Wikipedia Language Editions. The motivation of our research project is the drive for self-governance. We know that environmental and resource constraints could drive self-governance, so do heirs and lessons of community members. Therefore, we have seen a striking diversity as well as consonances and how different communities solve the same set of problems. It has been difficult for us to predict when a set of institutions will converge or diverge in terms of their solutions to the same governance challenge. That has led us to the research question. Will organizations facing a similar set of challenge build similar sets of rules? When they do converge on similar solutions, do they follow the same path to get there? How much does the policy development depend on websites, on culture, or anything else? Wikipedia has provided a great opportunity for us to study self-governance and collective action institutions. Wikipedia is a free online psychopedia with additions in 316 languages. All these language additions are hosted by a common umbrella organization, the Wikimedia Foundation, and they share a common physical server infrastructure. Each of the Wikipedia language additions is peer-produced and self-coveting. The users could write their own rules coded by languages. Wikipedia has been proved to be a productive platform for studying policy processes. The Wikipedia language additions are highly comparable collective action institutions. They are open community-driven structure, and they pursue the same mission under the same constraints. All these characteristics made Wikipedia language additions to be an ideal case for studying policy diffusion, which is the topic of our research. Using wiki data, especially media wikis across wiki linking page, we were able to construct the policy structure across Wikipedia language additions. On the right-hand side, we have the language policy bipartite network, where we have 60 shared policies and 245 language additions. The policies are the nodes in blue, and the language additions are the nodes in yellow. In total, there are 3,600 instances of policy adoptions. That's the edges in this network. First of all, let's take a look at the policy sharing patterns. On the right-hand side, we have the one mode projection of the policy language network. The nodes are wiki language additions. The edges are weighted by the number of shared policies. The network is very tightly connected, and in the core, there are 30 fully connected language additions sharing at least 37 policies with one another. The English addition has adopted nearly all the policies that are shared across the additions. We did not find no evidence of clustering based on cultural traditions, though. And secondly, we move on to looking at the drives for the order of adoption. This figure shows that popularity actually drives the order of adoption. On the x-axis, we have the order of policy creation within wiki, and on the y-axis, we have popularity rank. The figure shows that the more popular a policy is, the sooner it will be adopted within wikipedia and language addition. And certainly, we took a look at the sequence of policy adoption. A sequence clustering analysis shows that despite the variations, we have found a relatively strong policy adoption of sequence. That means we have found generally a common path for wikipedia language additions to adopt policies. This could be explained by shared constraints, for example, resource limits, shared omission, constitutional structure. A common policy adoption path suggests that these shared constraints may play a larger role than addition-specific constraints. In conclusion, wikipedia provides a rare case of highly comparable policy systems developing in parallel, which enables us to test the theories of policy development. Our analysis has found that policies are widely shared among different language additions. There's little evidence of policy adoption based on cultural traditions, and we have found that policy adoption is mainly driven by pop popularity. A sequence analysis reveals that we have a single developmental path across all these wikipedia language additions. I want to note that our analysis only look at shared policies. That means we could only answer questions like when a policy gets diffused, how it got diffused. We cannot answer questions like among all the policies, why some of them got diffused as others did not. To pursue the latter question, we would need to collect a more comprehensive set of policy, including not just the shared policies, but also policies that are unique to certain wikipedia additions. And secondly, I want to note that our analysis is based on the fact of policy. We didn't go into the content of the policy. It's very possible that even though wikipedia additions have the same policy under the same name, the rules under this policy could be completely contradictory. We would need to have future studies looking more into that. Certainly we were also wondering about the generalizability of policy adoption patterns we found out on wikipedia. How would that apply to, for example, policy adoption across political institutions. Would that be different? Would we still be seeing a common path right across different political institutions. And lastly, we have identified a single developmental ladder for policy adoption, but what are the mechanisms, right? Some of them are possible. The common mission of resource constraints or environmental factors and how and to what extent they could affect the policy development processes. Future studies would benefit from looking more into that. That's the end of my presentation. Thank you very much.