[Purpose/significance] This paper dissects the process and services of article association discovery in Connecting Repositories, and hopes to provide powerful reference for the recommendation and semantic linking of the content of articles in Chinese open access repositories. [Method/process] This paper analyzed the discovery process of article association based on semantic similarity and the semantic services based on article association. The discovery process of article association based on semantic similarity included metadata and full-text content harvesting, and semantic similarity calculation of article association. The semantic service based on the discovery process of article association included the CORE recommendation service and the linked open data service. And this paper summarized the application suggestions of CORE to Chinese institutional repositories. [Result/conclusion] This paper finds CORE system automatically harvests the metadata of the open access repositories through the existing OAI-PMH protocol, and further extracts the URI fields from the metadata to download the full-text through the HTTP protocol. Furtherly, providing article recommendation services and services of data linked articles based on the discovery of article semantic association enables third-party systems to utilize CORE datasets, it provides a powerful reference in recommendation and semantic linking of article association for open access repositories (such as institutional repositories and open access journals) in China.