Using Memetic Algorithm for Instance Coreference Resolution

TLDR

Instance coreference resolution is a key challenge in semantic web research, essential for web‑of‑data implementation and semantic data integration. This study proposes using a Memetic Algorithm to address instance coreference resolution by leveraging schema‑level matching results in a sequential framework. The authors formulate optimization models for schema‑ and instance‑level matching, define profile similarity measures and evaluation metrics assuming one‑to‑one alignment, and detail the Memetic Algorithm implementation. Experiments on OAEI benchmarks and real‑world datasets demonstrate that the proposed approach outperforms state‑of‑the‑art systems.

Abstract

Instance coreference resolution is an essential problem in studying semantic web, and it is also critical for the implementation of web of data and future integration and application of semantic data. In this paper, we propose to use Memetic Algorithm (MA) to solve this instance coreference problem in a sequential stage, i.e., the instance-level matching is carried out with the result of schema-level matching. We first give the optimization model for schema-level matching and instance-level matching. Then, we, respectively, present profile similarity measures and the rough evaluation metrics with the assumption that the golden alignment for both schema-level matching and instance-level matching is one-to-one. Furthermore, we give the details of the MA. Finally, the experiments of comparing our approach with the state-of-the-art systems on OAEI benchmarks and real-world datasets are conducted and the results demonstrate that our approach is effective.