Drug Evaluation Committee Model Prototype for Investigating the Use of AI in Data Management Operations
Data Science Expert Committee
June 2024
In the data management area, there are still many processes that rely on manual work, which not only burdens the workload of the person in charge, but also poses a risk of human error affecting data quality. Therefore, Task Force 1-1 of the Data Science Subcommittee in FY2023 decided to conduct three types of AI prototyping by examining familiar tasks in the data management area that could be made more efficient and less error-prone through the use of AI. In this prototyping, the members themselves carried out the entire process of selecting natural language processing models and large-scale language models to be used, preparing training and test data, and training, validating, and testing AI models to achieve the intended processing. This deliverable summarizes these prototyping efforts in the form of objectives, methods, results, and discussion. We hope that readers will read it as if they were "reliving" the prototyping process.
Japan Pharmaceutical Manufacturers Association Drug Evaluation Committee
Data Science Subcommittee FY2023 Task Force 1-1
Model Prototyping for Investigating the Use of AI in Data Management Operations