A methodology and a software for validating citation data and bibliographic metadata according to the OpenCitations Data Model

The thesis work introduces a methodology and a software for the validation of citation data and bibliographic metadata structured in tabular format, according to the OpenCitations Data Model (OCDM). In particular, the technical solution stems from the need of ensuring the validated data quality and its conformance to the data model by providing the users with machine-readable and
human-readable outputs both locating possible errors in the table and providing explanations of the reasons for validation failures. The two case studies chosen for implementing the methodology concern data ingestion in OpenCitations META and CROCI databases, respectively dealing with metadata of bibliographic entities and citation data. The validation procedure was structured on the variety of validation aspects to be handled, resulting in the following categorization: (1) validation of the table’s format with respect to its own format and model specifications; (2) validation of the ID values against the syntax of the corresponding schemes;
(3) verification of the IDs existence; (4) validation of semantic relations according to the OCDM, covering aspects not depending on the table format, such as the compatibility between ID and resource type for a given bibliographic entity. After an overall introduction to the thesis work, the first chapter presents a literature review on existing schema languages and validation software. The second chapter analyzes the case study and the features of the data involved, introducing the implementation methodology. The third and the fourth chapters describe the Python software of the validator respectively at high-level and in technical detail. The fifth chapter presents an evaluation of the software based on the process execution time, and it is followed by conclusive remarks on the work including an overview of possible future developments.