When LLMs Encounter Open-world Graph Learning: A Fresh View on Unlabeled Data Uncertainty

Recently, large language models (LLMs) have driven a systematic shift in the graph ML com- munity through the adoption of text-attributed graphs (TAGs). Although a variety of frameworks have been developed, most fail to properly ad- dress the challenge of data uncertainty in open- world environments. A representative source of such uncertainty is the limited availability of la- bels in large-scale datasets due to high annotation costs, where unlabeled nodes may belong to either known classes or novel, unknown classes. While node-level out-of-distribution detection and con- ventional open-world graph learning attempt to tackle this problem, two core limitations remain: ① Insufficient methods — existing approaches typically optimize semantics or topology in isola- tion for unknown-class rejection, failing to effec- tively integrate textual and structural information in TAGs; ② Incomplete pipelines — most stud- ies conduct only idealized analyses, such as as- suming a predefined number of unknown classes, which restricts practical utility for model updates and long-term deployment. To overcome these issues, we introduce the Open-world Graph Assis- tant (OGA), an LLM-based framework. OGA first performs unknown-class rejection via adaptive la- bel traceability (ALT), harmoniously combining semantic and topological cues, and then applies the graph label annotator (GLA) for unknown- class annotation, allowing unlabeled nodes to con- tribute to model training. In essence, OGA offers a new pipeline that fully automates the handling of unlabeled nodes in open-world environments, and we establish a systematic benchmark cover- ing four key aspects to validate its effectiveness and practicality through extensive experiments.

Large language models (LLMs) have helped computers better understand graphs that include text descriptions (like social networks or citation networks). However, a big challenge is that in real-world settings, many data points (nodes) have no labels — and worse, some belong to completely new, unknown categories. Most existing methods either ignore text and structure together or make unrealistic assumptions, like knowing exactly how many unknown categories exist. In this paper, the researchers introduce a new system called the Open-world Graph Assistant (OGA). OGA works in two steps: first, it decides which unlabeled data points are likely from known categories versus completely new ones by combining both the text meaning and the graph connections. Then, it automatically assigns useful labels to those new-category nodes so they can help train the model further. This creates a fully automated pipeline that works in messy, open-world environments without needing humans to predefine unknown classes. Extensive experiments show that OGA is effective across four different evaluation perspectives.