Precedent Influence Score

A quantitative metric computed over a citation graph that estimates the precedential weight of a legal case. It is derived from a node's in-degree centrality, the authority scores of citing cases, and temporal decay factors. Variants of PageRank and HITS algorithms are commonly adapted to account for the hierarchical nature of court systems, where citations from higher courts carry greater weight. Unlike a raw citation count, an authority score distinguishes between a case that is merely cited frequently and one that is cited by other highly authoritative decisions.

An NLP task that automatically categorizes the legal effect a citing case has on a cited authority. Standard labels include:

• Overruled: Explicitly invalidated
• Distinguished: Found inapplicable on facts
• Followed: Applied as controlling
• Criticized: Questioned but not overruled
• Harmonized: Reconciled with other authority This classification is the primary signal for determining whether a precedent remains good law and directly feeds into the sentiment weighting of edges in a Precedent Influence Score calculation.

A family of graph algorithms that iteratively distribute influence scores across a citation network. Starting from seed nodes or uniform distribution, these algorithms pass weight along directed citation edges. Key implementations include:

• Weighted PageRank: Incorporates treatment sentiment as edge weights
• HITS (Hubs and Authorities): Separates cases that cite many authorities from those that are cited by many hubs
• Personalized PageRank: Biases propagation toward a specific jurisdictional subset The resulting steady-state scores form the mathematical backbone of the Precedent Influence Score.

The polarity of a citing reference toward the cited authority, ranging from strongly supportive to strongly negative. This is a finer-grained signal than treatment type, capturing the rhetorical tone within a citation instance. A case may be 'followed' but with reluctant or critical language. Sentiment is typically extracted using fine-tuned transformer models trained on annotated legal corpora. In a Precedent Influence Score, negative sentiment edges act as dampening factors, reducing the flow of authority from a criticized decision to its citing cases.

The algorithmic identification of landmark decisions that originate major legal doctrines. Seminal cases exhibit a distinctive signature in the citation graph:

• High out-degree centrality: Cited by a vast number of subsequent cases
• Sustained citation velocity: Continued relevance over decades
• High betweenness centrality: Bridge between doctrinal clusters
• Low negative treatment ratio: Rarely overruled or criticized These cases serve as anchor nodes in authority propagation and often define the upper bound of the Precedent Influence Score scale.

The study of citation patterns over time to model how legal authority evolves. This analysis incorporates timestamps on citation edges to detect:

• Precedent aging: Declining citation frequency as doctrines mature
• Citation bursts: Sudden renewed interest in a dormant precedent
• Doctrinal shifts: Gradual replacement of one line of authority by another Temporal decay functions are essential to a dynamic Precedent Influence Score, ensuring that a case's influence reflects its current relevance rather than historical citation volume alone.