Understanding how attributes interact within networks is key to modeling complexity in social and informational systems. This work extends the MGJ model to estimate latent feature–feature interactions that drive or inhibit link formation in large graphs. By reformulating the problem as a perceptron-like learning task, it introduces scalable algorithms that infer structural regularities from high-dimensional relational data, bridging network analysis and machine learning foundations later applied to socio-technical behavior.

Collaborative knowledge systems evolve through messy, user-generated hierarchies. This study proposes a centrality-based pruning method that cleanses the Wikipedia category network by identifying structural redundancies and inconsistencies. By relying solely on endogenous information, it demonstrates how collective curation and algorithmic structure can be combined to improve the organization of open knowledge — work that later became the basis for a Wikipedia-based benchmark for graph neural networks, now widely used in research.

Discovering non-obvious relations in knowledge systems requires models that go beyond surface similarity. This paper presents an online margin-based learning algorithm that infers a latent matrix of category interactions to uncover hidden connections within Wikipedia’s hyperlink structure. The method efficiently scales to large graphs, revealing how semantic structures and user-generated organization jointly shape information discovery.

Networks emerge from both shared attributes and competition among their participants. This work introduces a latent-attribute generative model where connections depend on feature overlap and a node’s fitness to transmit attributes, allowing even new nodes to compete for links. The model reproduces key empirical regularities of real systems and provides a parsimonious explanation of growth and inequality in complex networks.