FOS: A Large-Scale Temporal Graph Benchmark for Scientific Interdisciplinary Link Prediction

TL;DR

FOS introduces a large-scale, time-aware benchmark for scientific interdisciplinarity by modeling yearly co-occurrence graphs of 65,027 fields across 19 domains, with semantic node embeddings to capture field meaning. Forecasting is formulated as a temporal link-prediction task to identify first-time field pairings, and a reproducible pipeline with multiple negative-sampling regimes is provided to evaluate state-of-the-art temporal GNNs. Through comprehensive experiments on the Art+Business subset, the study shows that no single model dominates across all regimes and that long historical context and rich semantic features, especially description embeddings, are crucial for predicting novel interdisciplinary links. The results align high-scoring predictions with later real-world publications, underscoring FOS's practical utility for surfacing emerging scientific directions; the dataset, splits, and code are released to advance research in forecasting scientific frontiers.

Abstract

Interdisciplinary scientific breakthroughs mostly emerge unexpectedly, and forecasting the formation of novel research fields remains a major challenge. We introduce FOS (Future Of Science), a comprehensive time-aware graph-based benchmark that reconstructs annual co-occurrence graphs of 65,027 research sub-fields (spanning 19 general domains) over the period 1827-2024. In these graphs, edges denote the co-occurrence of two fields in a single publication and are timestamped with the corresponding publication year. Nodes are enriched with semantic embeddings, and edges are characterized by temporal and topological descriptors. We formulate the prediction of new field-pair linkages as a temporal link-prediction task, emphasizing the "first-time" connections that signify pioneering interdisciplinary directions. Through extensive experiments, we evaluate a suite of state-of-the-art temporal graph architectures under multiple negative-sampling regimes and show that (i) embedding long-form textual descriptions of fields significantly boosts prediction accuracy, and (ii) distinct model classes excel under different evaluation settings. Case analyses show that top-ranked link predictions on FOS align with field pairings that emerge in subsequent years of academic publications. We publicly release FOS, along with its temporal data splits and evaluation code, to establish a reproducible benchmark for advancing research in predicting scientific frontiers.

Figures (9)

• Figure 1: Construction of the FOS Temporal Benchmark. Papers are mapped to concept nodes via the OpenAlex knowledge-graph hierarchy. Co-occurring concepts within a paper form concept pairs that define topological features, while each paper's publication year provides the temporal signal. Each concept node is represented as a semantic vector using an embedding model. These node embeddings, together with the topological and temporal descriptors, are aggregated into annual co-occurrence graphs that collectively constitute the FOS temporal benchmark.
• Figure 2: Impact of Historical Data Span on Model Performance. This plot illustrates AP and AUC-ROC scores, averaged across test years (2022-2024), for the DyGFormer model trained on varying historical data windows starting from 2002 to 2015 (with a fixed end year). Extending the historical span improves predictive performance, though with diminishing marginal gains as the starting year recedes further into the past.
• Figure 3: FOSart&business Node-Level Statistics (Part 1): Plots depicting (1) the number of active nodes per year, (2) the mean node degree per year, (3) the mean annual growth rate in node degrees relative to the prior year, and (4) the mean clustering coefficient per year.
• Figure 4: FOSart&business Node-Level Statistics (Part 2): Distributions of (1) node persistence spans (total active years) and (2) the number of nodes by year of last activity (recency).
• Figure 5: FOSart&business Edge-Level Statistics (Part 1): Plots showing (1) edge count and density over time and (2) the edge repetition rate (fraction of recurring edges) per year.