Eviction is arguably one of today’s most significant and pressing issues, affecting millions of households. On average, 3.6 million eviction filings occur annually across the United States. The eviction crisis is poised to intensify in the aftermath of the COVID-19 pandemic, displacing low-income families, restricting future housing options, and potentially leading to homelessness. A significant challenge in addressing this issue is the lack of systematically collected eviction data, with a vast amount of detailed information contained in unstructured court records, primarily in PDF format. Although filing data offer insight regarding this initial step in the eviction process, the events that follow remain largely understudied. This dissertation leveraged computational social science techniques, combining social and data science, to extract these unstructured data and analyze posteviction filing outcomes. This study explored three main issues: (a) the efficacy of computational methods in extracting information from unstructured court files; (b) the influence of individual, community, and macro-level factors on dismissal or judgments; and (c) the determinants of eviction by the sheriff following court judgment. The study utilized advanced document layout analysis based on natural language processing and computer vision to recognize information in PDF files and link this personal information to broader property, community, and county datasets. A classification model was used to identify important factors related to eviction filing outcomes. The analysis covered 56,070 unique cases derived from 772,629 PDF files spanning 2004–2022 in Pierce County, Washington, demonstrating high accuracy in data extraction (median Levenshtein similarity ratio of 1 and mean of 0.95). Key findings indicated that significant individual-level factors—such as race, property sale records, legal representation, taxable property value, and response to summons—influence eviction filing outcomes. At the community level, poverty rates and the proportion of rent-burdened households emerged as strong predictors. At the macro level, a housing price index and rent prices play a crucial role. The interaction between an individual’s race and the proportion of White people in the census tract shows that people of color experience different eviction filing outcomes compared to White individuals in the same community. The discussion touches on computational social science in eviction research and how variables at different levels affect eviction filing outcomes. The study findings have implications for social welfare interventions and policy, aiming to support affected families and mitigate the eviction crisis.