In this article, a proposed architecture for integrating knowledge graphs and a pipeline that automates the transformation of spreadsheet data into RDF triple knowledge graphs for the open data platform of the ELLAS research network is described. The main goal of the ELLAS research network project is to generate comparable data on gender gaps in STEM in Latin America, and since the data comes from various sources and formats, semantic web will be used to add meaning to the data. The architecture and pipeline proposal is to perform Extract, Transform, and Load (ETL) activities for data storage and availability to end users.

The underrepresentation of women in STEM fields needs to be highlighted through data to assist decisionmakers and public policy creators in addressing the issue effectively. However, the lack of structured, organized data published openly in this domain is still a reality. To address this problem, a Latin American research network called ELLAS was created. The project’s goal is to develop a platform with Semantic Web-based technologies to structure and concentrate data from Brazil, Peru, and Bolivia, initially. This paper presents the processes defined for the collection and curation of both unstructured and structured data, sourced from scientific articles, social networks, and existing open data. We explore the architecture design in a way that facilitates understanding of the details of the processes and the actors involved for each data source. We present the preliminary results fr...