Year: 2025
Type: conference-paper
Abstract: Currently, medicine manufacturing data is difficult to access and query as it is in a) unstructured format, ii) scattered at multiple locations in variety of formats iii) without meta-data and stored in files. To solve these problems, we have proposed a Digital Medicine Manufacturing - Extract-Transform-Load (DM2 ETL) tool to derive maximum value from the data acquisition effort to date and to allow future data to be integrated easily. DM2 ETL, with multiple components, is responsible for extraction, transformation and loading of heterogeneous medicine manufacturing data related to multiple instruments. Schema for experimental data in the medicine manufacturing domain has been designed that provides a structure for data and establishes linkage to meta-data. For central repository in structured format, we have used MS-SQL server. Once data is structured in MS-SQL, server, we have performed semantic integration of medicine manufacturing data using virtual knowledge graph. Medicine manufacturing data virtualisation is performed through integration of ontotext ontop platform. To achieve this task, we have developed Ontology Based Data Access (OBDA) mappings defined from DM2 database. OBDA file descriptors are developed to map relational schema to graphs. This helps us to access the medicine manufacturing data via SPARQL endpoint. This allows medicine manufacturing experts to access and query the overall data assets in an integrated way, by exploiting the semantics of the extracted information. As data is in interoperable format so its easy to connect it to other existing ontologies in the domain.
Authors: Naz, Tabbasum, Johnston, Blair
Language: English
Topic: Semantic Web and Ontologies
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7204014359