This work proposes to measure model complementarity through the performance of an oracle that, for each query, selects the best prediction among a considered set of models, hence providing an upper bound on the performance achievable through model combination.
Abstract
Knowledge graphs have become an important source of structured knowledge for Web applications, including search, question answering, and recommender systems. In these applications, link prediction can serve either as a prediction task itself or as a means to enrich incomplete knowledge graphs for downstream tasks. Interestingly, different link prediction models, or even different training runs of the same model, can produce substantially different predictions for the same query. This suggests a variability in the capture of the underlying knowledge by models, thus raising a fundamental question: to what extent do different models capture complementary knowledge, and how much of this knowledge could be recovered by combining them? We propose to measure model complementarity through the performance of an oracle that, for each query, selects the best prediction among a considered set of models, hence providing an upper bound on the performance achievable through model combination. Across several architectures and benchmarks, we find a substantial gap between individual models and their oracle, revealing that different models capture complementary knowledge. Yet, this complementarity rapidly saturates as more models are added, leaving a persistent subset of queries unsolved even by a large number of models. These findings reveal both the potential of model complementarity and a fundamental limit to what current link prediction models can collectively recover; thereby highlighting the need for further research to build robust Web applications.
Over the past two decades, Recommender Systems (RSs) have repeatedly changed how they integrate external knowledge, moving from constraints and side information to Knowledge Graphs (KG) and, more recently, Large Language Models (LLMs). This Past/Present/Future retrospective interprets that evolution as a sequence of re...
Claudio Pomo, Diego Baquero Sanz, Liam Claude Morris et al.· Proceedings of the 20th ACM...· 0 citations
Knowledge Graph Embedding models have been extensively used to learn representations of entities and relations in Knowledge Graphs for predicting missing links. However, the quality of the learned representations varies a lot across different areas of the graph. If previous research has loosely linked the problem to re...
Alberto Bernardi, Luca Costabello, Christophe Guéret· 0 citations
This work compares country-continent questions with noun, adjective, and code answers while keeping several fitted measurements distinct across Qwen, Llama, and Gemma to separate early readability, natural strength, causal steering, and later content dependence.
Wen-Lin Wei, Yuan Fang, Ren-He Jiang et al.· 0 citations
A neural regression model (LitEm) that enables transductive knowledge graph embedding models to predict numerical attributes within knowledge graphs and a co-training framework that jointly trains state-of-the-art transductive knowledge graph embedding models with LitEm, which improves link prediction performance mainl...
Rupesh Sapkota, Louis Mozart Kamdem Teyou, Moshood Yekini et al.· 0 citations
This work introduces EvoSkillRec, a promotion-and-reuse framework for cumulative recommender architecture evolution that decomposes recommenders into atomic executable skills and represents architectures as typed skill genomes, with each skill equipped with input--output types, semantic annotations, and implementation...
Xiao-Peng Li, Kuo Cai, Bo Chen et al.· arXiv.org· 1 citation
This paper systematically evaluates the overall performance of representative CS approaches, investigates the impact of intrinsic properties of ground-truth communities, and analyzes the approaches' generalizability within a unified experimental framework to advocate generalizability as a pivotal direction for future r...
Yu-Han Zhou, Zheng Wu, Qing Liu et al.· Proceedings of the ACM on Ma...· 0 citations
Related blog posts
Microsoft Research Blog· microsoft.comNov 16, 2026
AI coding agents are increasingly help developers fix bugs and build features largely on their own, but every interaction with their models consumes tokens. In each session, an agent sends instructions, tool descriptions, and session history to its model. Most efforts to reduce this cost focus on individual tasks. Yet developers return to the same […] The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduNov 13, 2026