DGRA e.V. - STUDIENARBEITEN / MASTER-THESIS

Summary

Language: English

The complexity of search requests for regulatory professionals is ever growing and at the same time the volume of regulatory documents and information is rising exponentially. The primary idea within this Master Thesis is to create a regulatory dictionary in order to ensure easy searchability on regulatory websites, to facilitate the retrieval of regulatory information and to make the dictionary accessible for MDRA students. During the project it became clear that not only a regulatory dictionary could be developed, but moreover with the help of the Text Mining method it would be possible to generate an extraction system for regulatory information by building a regulatory ontology.

The challenge of Text Mining lies in deriving linguistically expressed information from free text documents for automatic analysis. This method facilitates to cope with the overwhelming amount of text, especially in the Life Science field, e.g. Regulatory Affairs.

To this day it is only possible to research for regulatory information on authorities websites via a key word query, comparable to a „Google query“. As result of the query one get several text documents which have to be perused in particular for gaining the requested information.

The development target of the current pilot project is to design a retrieval technology (information extraction system) which allows complex queries. The design of tailored Text Mining approaches for indexing, structuring and extraction of relevant information from authorities websites could strongly support drug development.
This Master Thesis describes all taken process-steps in order to demonstrate via a pilot-run how Text Mining would work in the domain of Regulatory Affairs, tries to give a definition of Regulatory Text Mining and introduces in the Text Mining technology. The built regulatory ontology which relates to terms and definitions of the Centralised Procedure is explained and illustrated in the Annex of the Thesis. Subsequently in the chapter Field of Application it is demonstrated how Regulatory Text Mining could be used for research in the regulatory domain. In the end an outlook is given about the current status of the project, possible future work and new projects in this area.