This chapter introduces the network machine learning landscape, bridging traditional machine learning with network-specific approaches. It defines networks, contrasts them with tabular data structures, and explains their ubiquity in various domains. The chapter outlines different types of network learning systems, including single vs. multiple network, attributed vs. non-attributed, and model-based vs. non-model-based approaches. It also discusses the scope of network analysis, from individual edges to entire networks. The chapter concludes by addressing key challenges in network machine learning, such as imperfect observations, partial network visibility, and sample limitations. Throughout, it emphasizes the importance of statistical learning in generalizing findings from network samples to broader populations, setting the stage for more advanced concepts in subsequent chapters.

Neil Steinkamp and Samantha DiDimenico, strategic consultants who have done extensive work on access-to-justice issues, offer a unique how-to guide for engaging courts and community stakeholders in order to generate quantitative and qualitative data that can contribute to reform efforts. Focusing on “civil Gideon,” a growing set of efforts to establish a “right to counsel” akin to what criminal defendants have long enjoyed under the Sixth Amendment, Steinkamp offers a step-by-step roadmap for developing an empirically rigorous and comprehensively informed dialogue toward regulatory reform.

Judge Carolyn Kuhl (L.A. Superior Court), until recently the chief judge of the nation’s largest trial court system, offers an important contribution to the debate about whether and how to relax “courthouse UPL” – the possibility that judges, court clerks, other court staff, and AI-enabled chatbots might plausibly narrow the justice gap by providing self-represented litigants with necessary assistance. At once a history lesson and an in-the-trenches look at a decade of L.A. court reforms, Judge Kuhl shows how the anxieties about judicial and court neutrality have given way to a rich array of reform options that are producing concrete lessons for other judicial reformers looking for alternatives to conventional forms of legal help.

This chapter presents a framework for learning useful representations, or embeddings, of networks. Building on the statistical models from Chapter 4, we explore techniques to transform complex network data into vector representations suitable for traditional machine learning algorithms. We begin with maximum likelihood estimation for simple network models, then motivate the need for network embeddings by contrasting network dependencies with typical machine learning independence assumptions. We progress through spectral embedding methods, introducing adjacency spectral embedding (ASE) for learning latent position representations from adjacency matrices, and Laplacian spectral embedding (LSE) as an alternative approach effective for networks with degree heterogeneities. The chapter then extends to multiple network representations, exploring parallel techniques like omnibus embedding (OMNI) and fused methods such as multiple adjacency spectral embedding (MASE). We conclude by addressing the estimation of appropriate latent dimensions for embeddings. Throughout, we emphasize practical applications with code examples and visualizations. This unified framework for network embedding enables the application of various machine learning algorithms to network analysis tasks, bridging complex network structures and traditional data analysis techniques.

This appendix provides a concise introduction to key machine learning techniques employed throughout the book. It focuses on two main areas: unsupervised learning and Bayesian classification. The appendix begins with an exploration of K-means clustering, a fundamental unsupervised learning algorithm, demonstrating its application to network community detection. It then discusses methods for evaluating unsupervised learning techniques, including confusion matrices and the adjusted Rand index. The silhouette score is introduced as a metric for assessing clustering quality across different numbers of clusters. The appendix concludes with an explanation of the Bayes plugin classifier, a simple yet effective tool for network classification tasks.

Rebecca Aviel (Denver University (Sturm) Law) draws on her deep expertise in family law to illuminate ways in which domestic relations cases are exceptional relative to other legal areas where access concerns are acute. Family law’s exceptionalism, she contends, justifies thoroughgoing changes to that system’s adversarial architecture, such as permitting a single lawyer to represent both sides in a divorce, that are well-tailored to family law even if nonstarters in other parts of the civil justice system. Aviel also suggests that some innovative family law programs might travel well, informing reforms in other civil justice contexts even where they cannot be directly replicated.

Allison Hoffman (University of Pennsylvania Carey Law), an expert on health care regulation, focuses on tectonic changes to health care in recent decades. She offers a bracing account of these shifts, arguing that American doctors may have overreached in their efforts at influencing health care regulation. In so doing, physicians created profit pools that corporate interests proved all too adept at capturing, leaving doctors with lower professional status than they might have otherwise enjoyed. Hoffman suggests that lawyers, and legal reformers more generally, might learn from physicians’ cautionary tale of protectionism and profit.

Philip G. Peters, Jr. (University of Missouri Law) examines whether nurse practitioners (NPs) and physician assistants (PAs) offer a promising template for limited license legal professionals. He interrogates the rise of these professions in the medical field, asking, among other things: Do they deliver quality services despite training that is significantly shorter and less expensive than the training of physicians? Do they reduce consumer costs? And do they improve access to care for underserved populations? The chapter also outlines the strategic factors underlying the remarkable success that the NP and PA professions have had, at least until recently, in statehouses across the country and then notes the arguments being made now by physicians against freeing NPs and PAs from all physician oversight. The chapter ends by identifying key lessons from this history for those seeking to create new categories of limited license legal professionals.

Genevieve Lakier (University of Chicago Law) examines Upsolve v. James, where a district court enjoined the application of New York state’s unauthorized practice of law statutes to the Justice Advocates that the nonprofit organization, Upsolve, planned to train, to assist low-income New Yorkers file for bankruptcy. The opinion represents a clear victory for the access-to-justice movement. But it also represents a potentially significant change in how courts understand the First Amendment to apply in unauthorized-practice-of-law cases. Although the decision may be overturned on appeal, the logic of the opinion thus makes clear the promise that what critics have sometimes described as a “Lochnerized” First Amendment holds out to access-to-justice advocates, as well as some of its perils. In this chapter, Lakier explains why the decision is significant, embeds it within a broader story of doctrinal transformation, and spells out some of the benefits and costs of using a Lochner-like First Amendment to promote access to justice.