diff --git a/CNAME b/CNAME index 2c37500a7ce..03c0553dc40 100644 --- a/CNAME +++ b/CNAME @@ -1 +1 @@ -jayrobwilliams.com \ No newline at end of file +missa7481.github.io diff --git a/_config.yml b/_config.yml index 0242a60636e..b838e54911e 100644 --- a/_config.yml +++ b/_config.yml @@ -7,13 +7,13 @@ # Site Settings locale : "en-US" -title : "Rob Williams" +title : "Zihui Ma" title_separator : "-" -name : &name "Rob Williams" -description : &description "Postdoc in Political Science" -url : https://jayrobwilliams.com # the base hostname & protocol for your site e.g. "https://mmistakes.github.io" +name : &name "Zihui Ma" +description : &description "Phd @ UMD" +url : https://missa7481.github.io # the base hostname & protocol for your site e.g. "https://mmistakes.github.io" baseurl : "" # the subpath of your site, e.g. "/blog" -repository : "jayrobwilliams/jayrobwilliams.github.io" +repository : "missa7481/missa7481.github.io" teaser : # filename of teaser fallback teaser image placed in /images/, .e.g. "500x300.png" breadcrumbs : false # true, false (default) words_per_minute : 160 @@ -57,7 +57,7 @@ yandex_site_verification : # Social Sharing mastodon: - username : "fosstodon.org/@jayrobw" + username : #"fosstodon.org/@jayrobw" twitter: username : facebook: @@ -84,16 +84,16 @@ analytics: # Site Author author: - name : "Rob Williams" + name : "Zihui Ma" avatar : "profile.png" - bio : "Data Scientist" + bio : "PhD @ UMD" location : employer : pubmed : - orcid : "http://orcid.org/0000-0001-9259-3883" - googlescholar : "https://scholar.google.com/citations?user=fiaPSmgAAAAJ" - email : "rob.williams@wustl.edu" - researchgate : # example: "https://www.researchgate.net/profile/yourprofile" + orcid : "http://orcid.org/0000-0002-2836-280X" + googlescholar : "https://scholar.google.com/citations?user=-lWZw2IAAAAJ&hl=en" + email : "zma88@umd.edu" + researchgate : "https://www.researchgate.net/profile/Zihui-Ma-3" uri : bitbucket : codepen : @@ -101,15 +101,15 @@ author: flickr : facebook : foursquare : - github : "jayrobwilliams" - gitlab : "jayrobwilliams" + github : "missa7481" + gitlab : #"jayrobwilliams" google_plus : keybase : instagram : impactstory : #"https://profiles.impactstory.org/u/xxxx-xxxx-xxxx-xxxx" lastfm : - linkedin : - mastodon : "fosstodon.org/@jayrobw" + linkedin : "linkedin.com/in/zihui-helen-ma-b77382a4" + mastodon : #"fosstodon.org/@jayrobw" pinterest : soundcloud : stackoverflow : #"https://stackoverflow.com/users/10912314/jayrobwilliams" diff --git a/_data/navigation.yml b/_data/navigation.yml index ad181de2ff9..38ea77f5381 100644 --- a/_data/navigation.yml +++ b/_data/navigation.yml @@ -6,14 +6,14 @@ main: - title: "Research" url: /research/ + - title: "Research2" + url: /research/ + - title: "Teaching" url: /teaching/ - - title: "Software" - url: /software/ - - - title: "Posts" - url: /posts/ + - title: "Talks" + url: /talks/ - title: "CV" url: /cv/ diff --git a/_pages/about.md b/_pages/about.md index 1ed88d879e4..e4694e86889 100644 --- a/_pages/about.md +++ b/_pages/about.md @@ -8,23 +8,11 @@ redirect_from: - /about.html --- -Welcome! I am a data scientist applying machine learning tools and causal -inference techniques to remote sensing data. I am also an affiliated -researcher with the [Data-driven Analysis of Peace Project](https://dapp-lab.org) -and a research collaborator with the -[Research on International Policy Implementation Lab](https://bridgingthegapproject.org/ripil). +Greeting! I am a Ph.D. candidate in the deparment of civil and environmental engineering (CEE) at the University of Maryland (UMD), currently conducting reserach as part of [Dr.Baecher's](https://cee.umd.edu/clark/faculty/244/Gregory-B-Baecher) esteemed reserach group. I am also a research collaborator with the +[Risk-Informed Solutions in Engineering Laboratory](https://riselab.umd.edu/). I received M.S.in civil engineering in project management from UMD in 2020 and M.S.in civil engineering in structure engineering from San Franciso State University (SFSU) in 2018. + +My academic work has been published or is forthcoming in esteemed journals such as the *International Journal of Disaster Risk Reduction*, the *Journal of Biomedical Informatics*, the *Journal of Construction Engineering and Management*, the *Fire Safety Journal*, and the J*ournal of Air Transport Management*, among others. These [researches](reserach) focus on harnessing crowdsourced information to enhance informed decision-making in emergency management spanning a spectrum from disaster response and pandemic preparedness to humanitarian crises. The ultimate goal is to play a role in fostering resilient and sustainable communities, thus contributing to a better future for all. + + + -I earned my Ph.D in Political Science from the -[University *of* North Carolina *at* Chapel Hill](https://www.unc.edu) and my -B.A. in Political Science from [Haverford College](https://www.haverford.edu). -My academic work has been [published](publications) or is forthcoming in -*International Studies Quarterly*, *Conflict Management and Peace Science*, -*Political Science Research and Methods*, and *PS: Political Science & Politics*, -among other outlets. This [research](research) explores the causes and -consequences of political violence using a broad variety of methods such as -latent variable models, geospatial analysis, and big data. While primarily -focused on civil conflict, it also examines contentious political phenomena -including terrorism and economic statecraft, and develops new measures of -institutions in international relations. I have [teaching](teaching) experience -in both quantitative methodology and international relations, and am a certified -instructor with [The Carpentries](https://carpentries.org). diff --git a/_pages/cv.md b/_pages/cv.md index 252eaceb068..9e24b2b3d96 100644 --- a/_pages/cv.md +++ b/_pages/cv.md @@ -7,6 +7,6 @@ redirect_from: - /resume --- - + -You can download a PDF copy of my CV [here](/files/pdf/Williams CV.pdf). +You can download a PDF copy of my CV [here](/files/pdf/Zihui Ma_CV2023.pdf). diff --git a/_pages/research.md b/_pages/research.md index 9945ecb4248..b0a2c317fed 100644 --- a/_pages/research.md +++ b/_pages/research.md @@ -7,27 +7,15 @@ header: og_image: "research/ecdf.png" --- -My academic research falls into two main areas: understanding the influence of -geography on actor behavior before, during, and after civil conflict, and -developing new tools to improve the study of institutions (both formal and -informal) in peace and conflict. One strand of research in this first area -explores how the territories that ethnic groups inhabit shape rebel group -formation and condition their relationship with the state. My interest in -geography also informs projects on active conflicts including the targeting of -UN peacekeepers by insurgent groups, civilian victimization after rebel -territorial conquest, and communal violence in fragile settings. +My primary research thrust revolves around the effective utilization of human-generated data, with a specific +emphasis on harnessing crowdsourced information. My objective is twofold: firstly, to advance our comprehension +of community resilience and to facilitate informed decision-making and management strategies for +natural disasters and unforeseen events. Secondly, to explore the nexus of humans, infrastructure, and climate change in urban environments, +with a vision of contributing to a sustainable future. At the core of my research lies a robust data-driven +framework, encompassing cutting-edge techniques such as machine learning, deep learning, and epidemic +modeling. My research interest extends across several pivotal domains: computational simulations for infrastructure +reliability, AI-enabled risk assessment and monitoring, and transforming education with AI. -My other main research agenda uses advanced methods to develop new measures of -institutions. One project uses Bayesian item response theory to measure the -strength of peace agreements as a latent variable and free researchers from -post-treatment bias caused by using the duration of agreements as a proxy for -their strength. In others, I apply unsupervised learning techniques to over a -billion observations of product-level international trade data to measure -economic interdependence and illicit economic exchange. - -In a new avenue of research, I leverage social media data to explore -participation in extremist movements across multiple contexts, gaining insight -into the early stages of radicalization. diff --git a/_pages/teaching.md b/_pages/teaching.md index 603c56b840f..9fe1e0d36a4 100644 --- a/_pages/teaching.md +++ b/_pages/teaching.md @@ -3,38 +3,22 @@ permalink: /teaching/ title: "Teaching" --- -Research plays a central role in my teaching as students improve their -analytical skills and master the tools of data analysis through hands-on -experience. I have taught undergraduate courses on political violence and -statistical methodology. While at UNC, I taught the graduate statistics lab for -Advanced Topics in Political Data Science, where my work was recognized by the -Political Science Department's Earle Wallace Award for Graduate Student -Teaching. I also served as a teaching assistant for courses in international -relations and American politics at UNC, in addition to the ICPSR Summer Program -where I was a teaching assistant for a course on Bayesian modeling in the social -sciences. I am also a -[certified instructor](https://carpentries.org/instructors/#jayrobwilliams) with -[The Carpentries](https://carpentries.org/), which develops evidence-based -methods for teaching "essential data and computational skills for conducting -efficient, open, and reproducible research." +My journey through teaching and mentoring has been a profound and evolving experience. +## Teaching Assistant (mentored over 300 studnets) +- Project Cost Accounting and Finance (ENCE661) +- Introduction to Project Management (ENCE320) +- Introduction to Construction Management (ENCE325) +- Legal Aspects of Architectural and Engineering Practice (ENCE425) -You can view my teaching portfolio [here](/files/pdf/teaching/Portfolio.pdf). -You can find a selection of my teaching materials, including all of the labs -from Advanced Topics in Political Data Science, [here](/teaching-materials). +## Course Designer -## Washington University in St. Louis -- Pol Sci 3090: The Scientific Study of Civil War (Spring 2020) - - [Syllabus](/files/pdf/teaching/PS 3090 Syllabus.pdf) -- Pol Sci 3171: International Conflict Management & Resolution (Fall 2019) - - [Syllabus](/files/pdf/teaching/PS 3171 Syllabus.pdf) +Edx course: [Developing the Risk Management Plan with Expert Judgement](https://www.edx.org/learn/engineering/the-university-of-maryland-college-park-developing-the-risk-management-plan-with-expert-judgement). -## The University of North Carolina at Chapel Hill -- Poli 281: Quantitative Research in Political Science (Spring 2019) - - [Syllabus](/files/pdf/teaching/POLI 281 Syllabus.pdf) -- Poli 891: Lab for Advanced Topics in Political Data Science (Fall 2017, Fall 2018) - - [Syllabus](/files/pdf/teaching/POLI 891 Syllabus.pdf) +## Mentor +- Topic 1: "Impact of hurricanes on healthcare facilities" for one graduate student +- Topic 2: "The application of natural language processing in nature disaster” for one undergraduate studnet +- Topic 3: "Misinformation in pandemic" for one undergraduate student -## ICPSR Summer Program -- Introduction to Applied Bayesian Modeling (Summer 2017) - - [Syllabus](/files/pdf/teaching/bayes2017.pdf) + +Central to my teaching and mentoring philosophy is the commitment to fostering inclusive classroom environments that actively embrace diversity and encourage open dialogue. I am dedicated to ensuring that every student, regardless of their background, feels empowered to express their ideas and participate in critical discussions. diff --git a/_research/conflict-preemption.md b/_research/conflict-preemption.md deleted file mode 100644 index e150a519248..00000000000 --- a/_research/conflict-preemption.md +++ /dev/null @@ -1,21 +0,0 @@ ---- -title: "Conflict preemption" -layout: single-portfolio -excerpt: "" -collection: research -order_number: 10 -header: - og_image: "research/epr.png" ---- - -In this research I ask why some rebel groups fight for secession and independence, while others are willing to use violence to secure more autonomy and self-governance within an existing state. I argue that because rebel groups are strategic actors, they realize that military victory or plebiscite is not the end of their political struggle; if they gain independence, they must then create a new state. States are territorial entities, and so the trajectory of any new state will be greatly influenced by the resources and challenges its territory holds. Knowing this, rebel groups whose territory is more conducive to governance and administration will push for independence, while groups whose territory is less suited will fight for autonomy within the state. However, governments are aware of which groups inhabit territories most suitable to secession and employ various measures to try and stop these conflicts before they can begin, such as China’s pervasive electronic tracking of Uyghur citizens in Xinjiang. - -To test these arguments, I focus on rebel movements tied to ethnic groups with defined homelands. By doing so, I am able to exploit geospatial data on population and government activity to compare the governability of subnational territories cross-nationally. I also explore the temporal dynamics of these phenomena in smaller scale studies focusing on different government preemption tactics. I explore how states actually conduct this surveillance of their populations, using an agent-based model to predict when these monitoring efforts are likely to fail. Such simulation based approaches help us understand the role that information and communication technology can play in helping governments main control of restive regions. - -## Article - -Rob Williams. "Turning the Lights on to Keep Them in the Fold: How Governments Preempt Secession Attempts." *Conflict management and Peace Science*. - -> There are many regions that meet the necessary conditions for sovereign governance in the world, but few secessionist conflicts. I argue that this relative paucity of secessionist violence is the result of government preemption of potential secessionist movements. Using cross-national geospatial data from 1992 to 2013, I find that governments invest more, measured via nighttime light emissions, in more secession-prone regions. The same factors that make territory attractive for secession, such as large populations and international borders, also make governments willing to work to retain control of that territory, contributing to the scarcity of separatist civil conflicts. - -[Article](https://doi.org/10.1177/07388942211015242){: .btn--research} [Preprint](/files/pdf/research/Turning the Lights on.pdf){: .btn--research} [Supplemental Information](/files/pdf/research/Turning the Lights on SI.pdf){: .btn--research} [Replication Archive](https://journals.sagepub.com/doi/suppl/10.1177/07388942211015242){: .btn--research} [GitHub Repo](https://github.com/jayrobwilliams/conflict-preemption){: .btn--research} [Poster](/files/pdf/research/PSS 2018 Poster.pdf){: .btn--research} diff --git a/_research/event-data.md b/_research/event-data.md index 5c7c5bffcca..f12ff5b421b 100644 --- a/_research/event-data.md +++ b/_research/event-data.md @@ -1,5 +1,5 @@ --- -title: "Conflict event data" +title: "Blackout Community Resilience Assessment" layout: single-portfolio excerpt: "" collection: research @@ -8,26 +8,12 @@ header: og_image: "research/map.png" --- -In this set of projects, I leverage geospatial event data to explore the microlevel dynamics of political violence. What role can violence at the local level serve in advancing broader political ends? What explains patterns of action and reaction between actors engaged in different types of political violence? How do the differing roles the UN peacekeepers play affect the likelihood that rebel fighters will target them? +The public’s responses to sudden power outages offer valuable insights into a community’s ability to adapt and recover from crises. In this project, we proposed a framework to analyze sentiment and behavioral patterns, assessing community resilience during the 2019 New York City blackout. Through the identification of six distinct behavioral types (including “seek information,” “adjust schedules,” “find shelter,” “enjoy the situation,” “engage in altruistic acts,” and “report incidents”), we introduced an index to track shifts in public responses over time. Our analysis of these trending indexes revealed that New York City residents rapidly adjusted their routines, suggesting a robust community resilience in the face of the power outage. This study not only advances the methodologies employed in emergency management research but also contributes to our understanding of community resilience during emergency events. ## Article -Christian Oswald, Melanie Sauter, Sigrid Weber, and Rob Williams. "Under the Roof of Rebels: Civilian Targeting After Territorial Takeover in Sierra Leone." *International Studies Quarterly*. +Li, L., Ma, Z., & Cao, T. (2020). "Leveraging social media data to study the community resilience of New York City to 2019 power outage." *International Journal of Disaster Risk Reduction*. -> Do rebels target civilians as part of the process of establishing control in their territories? This research note shows that transition periods after rebels gain territorial control are remarkably violent for civilians. Speaking to the civilian victimization and rebel governance literature, we investigate the immediate time period after rebels successfully capture and hold territory. We argue that rebels use violence to gain compliance in newly captured territories until they are able to build up local capacities and institutions for peaceful governance. To test this argument, we draw on methodological advances in integrating event data and combine multiple datasets to study patterns of violence perpetrated by the Revolutionary United Front in Sierra Leone from 1997-2001. The findings of our spatiotemporal analysis show that civilian targeting increases in the period after rebels capture territory from the government compared to areas without territorial takeover, suggesting that life under the roof of rebels is initially more dangerous for civilians. +> Power outages across the world have severe social impacts. The public's responses to power outages provide valuable insights into their capacities of adapting crisis and an invaluable perspective to demonstrate community resilience. As social media has connected people in the community, the discussion on social media can reflect their responses as a criterion of resilience throughout power outages in a timely and effective manner. In the field of power outages, the potentials of social media data have only been investigated with recent advancements of big data techniques. Nonetheless, studies focusing on community resilience using social media data are quite limited. We filled this gap by introducing a novel and quantitative method to study the community resilience throughout power outages, based on a case study on the Manhattan blackout occurred in July 2019. The study examined community resilience from both mental and behavioral perspectives via sentiment analysis and behavior analysis. The sentiment analysis was used to track people's mental outlook and shape the overall mental status during and after this emergency. On the other side, six major patterns of behaviors were identified, and the behavioral index was defined to learn how the community responded to power outages amid the response and recovery periods. Both the mental and behavioral results reveal that New York City recovered at approximately one and a half hours after the blackout occurred, implying a strong community resilience to such short and emergent power outage events. -[Article](https://doi.org/10.1093/isq/sqaa009){: .btn--research} [Preprint](/files/pdf/research/Under the Roof of Rebels.pdf){: .btn--research} [Supplemental Information](/files/pdf/research/Under the Roof of Rebels SI.pdf){: .btn--research} [Replication Archive](https://doi.org/10.7910/DVN/BEKPWV){: .btn--research} - -## Working papers - -William G. Nomikos, İpek Ece Şener, and Rob Williams. "Does UN Peacekeeping Protect Civilians? Evidence from the Border between Burkina Faso and Mali." - -> Research in political science has shown that UN peacekeeping operations are an important tool for ending civil war violence. However, much less is known about how UN peacekeepers affect civilian victimization. Given that civilians bear the primary costs of intrastate conflict, understanding how international actors can contribute to the resolution of violence affecting them is a pressing concern. How does the presence of UN peacekeepers affect civilian victimization? We address this question by offering a straightforward empirical test of how UN peacekeeping patrols affect the likelihood that there will be violence against civilians. We build on the existing literature and established practices of peacekeeping to argue that peacekeepers deter violence against violence. To test our argument, we examine the case of Mail, the site of large-scale communal violence managed by UN peacekeepers since 2013. We employ a Geographic Regression Discontinuity Design (GRDD) around the border of Mali and Burkina Faso to estimate the causal effect of deploying peacekeepers to an area with growing communal tensions. Ultimately, our research provides robust causal evidence that UN peacekeeping works at the local level to protect civilians. - -[Working Paper](https://osf.io/preprints/socarxiv/5jmq4/){: .btn--research} - -Patrick Hunnicutt, William G. Nomikos, and Rob Williams. "Non-Combatants or Counter-Insurgents? The Strategic Logic of Violence against UN Peacekeeping." Presented at the Annual Conference of the American Political Science Association, San Francisco, CA, September, 2020. - -> Despite the wealth of academic research on United Nations (UN) peacekeeping operations, we know remarkably little about the causes of violence against peacekeepers. The dramatic increase in peacekeeper casualties over the past decade make this omission particularly problematic. This article demonstrates that violence against peacekeepers stems from strategic motivations. Peacekeepers in multidimensional PKOs serve as substitute providers of governance and security, working to bolster perceived state capacity and legitimacy in areas where the government cannot send its own forces. Insurgents target peacekeepers in expectation of a PKO unit’s capacity to win over the support of local civilians. We argue that insurgents rely on three primary heuristics to predict the downstream efficacy of peacekeeping forces: personnel composition, peacekeeper nationality, and local levels of insurgent control. We test our theory using an original dataset of geocoded UN multidimensional peacekeeping deployments peacekeeping deployments. Using primary documents sourced directly from the UN covering 10 multidimensional peacekeeping operations from 1999-2018, we present comprehensive time-series data on UN peacekeeper deployment location. We find preliminary evidence that peacekeepers are targeted because of their cultural similarity with noncombatants and, in some cases, because they patrol areas where insurgents have political control. - -[Working Paper](https://osf.io/ta96y/){: .btn--research} +[Article](https://doi.org/10.1016/j.ijdrr.2020.101776){: .btn--research} diff --git a/_research/sa.md b/_research/sa.md new file mode 100644 index 00000000000..a73af86e208 --- /dev/null +++ b/_research/sa.md @@ -0,0 +1,19 @@ +--- +title: "Wildfire SA evaluation" +layout: single-portfolio +excerpt: "" +collection: research +order_number: 10 +header: + og_image: "research/epr.png" +--- + +This research focus on leveraging social media data (i.e. Twitter) to comprehensively assess public dynamic situational awareness during 2020 wildfire season at the city-level. In this study, I employed Bidirection-al Encoder Representations from Transformers (BERT) topic modeling to cluster Twitter data and conducted a temporal-spatial analysis to understand topic distribution across various regions. Additionally, I integrated the Susceptible-Infected-Recovered (SIR) model to quantitatively measure the extent and speed of topic diffusion, facilitating more precise resource allocation. The results of the temporal-spatial analysis highlighted a close alignment between topic diffusion and wildfire locations and timelines, showcasing the real-time nature of Twitter discussions in response to unfolding events. Moreover, the findings from the topic-based SIR model revealed that the pace of topic diffusion corresponded with wildfire propagation patterns, reflecting varying levels of public awareness and responses. Furthermore, the study underscored the diversity of concerns expressed by different communities, emphasizing the need to tailor disaster responses to meet local needs effectively. + +## Article + +Ma, Z.*, Li, L., Hemphill, L., & Baecher, G. B. (2023). "Investigating disaster response through social media data and the Susceptible-Infected-Recovered (SIR) model: A case study of 2020 Western U.S. wildfire season." + +> Effective disaster response is critical for affected communities. Responders and decision-makers would benefit from reliable, timely measures of the issues impacting their communities during a disaster, and social media offers a potentially rich data source. Social media can reflect public concerns and demands during a disaster, offering valuable insights for decision-makers to understand evolving situations and optimize resource allocation. We used Bidirectional Encoder Representations from Transformers (BERT) topic modeling to cluster topics from Twitter data. Then, we conducted a temporal-spatial analysis to examine the distribution of these topics across different regions during the 2020 western U.S. wildfire season. Our results show that Twitter users mainly focused on three topics:"health impact," "damage," and "evacuation." We used the Susceptible-Infected-Recovered (SIR) theory to explore the magnitude and velocity of topic diffusion on Twitter. The results displayed a clear relationship between topic trends and wildfire propagation patterns. The estimated parameters obtained from the SIR model in selected cities revealed that residents exhibited a high level of several concerns during the wildfire. Our study details how the SIR model and topic modeling using social media data can provide decision-makers with a quantitative approach to measure disaster response and support their decision-making processes. + +[Preprint](https://doi.org/10.48550/arXiv.2308.05281){: .btn--research} diff --git a/files/pdf/Zihui Ma_CV2023.pdf b/files/pdf/Zihui Ma_CV2023.pdf new file mode 100644 index 00000000000..3df7336ad3b Binary files /dev/null and b/files/pdf/Zihui Ma_CV2023.pdf differ diff --git a/images/profile.png b/images/profile.png index fe2801145fe..0f28eba050f 100644 Binary files a/images/profile.png and b/images/profile.png differ diff --git a/images/research/nyc.png b/images/research/nyc.png new file mode 100644 index 00000000000..2c751f8cc7a Binary files /dev/null and b/images/research/nyc.png differ diff --git a/images/research/sir.png b/images/research/sir.png new file mode 100644 index 00000000000..d18e8984610 Binary files /dev/null and b/images/research/sir.png differ