Repository hosted by TU Delft Library

Home · Contact · About · Disclaimer ·
 

Identification of new and emerging occupational risks using a text mining based information system

Publication files not online:

Author: Pronk, A. · Goede, H. · Lucas Luijckx, N. · Brug, F. van de · Cnossen, H. · Tielemans, E.
Type:article
Date:2011
Place: Amsterdam
Source:Proceedings of the International Congress on Tracing New Occupational Diseases: methodology, recent findings and implications for OHS-policy, 7-8 April 2011, Amsterdam, The Netherlands, 39
Identifier: 462835
Keywords: Informatics · Future Internet Use · Information Society · Life · QS - Quality & Safety · EELS - Earth, Environmental and Life Sciences

Abstract

Introduction On the internet and in scientific databases relevant information is available on new and emerging occupational risks. However, the amount of information is enormous and the information is scattered over multiple and diverse data sources complicating the full utilization of the data sources. TNO has developed an innovative information system that can filter relevant knowledge on food risks from the growing flow of information. This system will be used as a basis for the development of a similar system for occupational risks. Methods The functionality of the method is based on the combination of advanced text mining technologies, expert knowledge and the application of up-to-date scientific databases. Information on emerging hazards can be obtained by defining tailor-made search queries, for which information (in e.g. ontologies) on occupational hygiene, substances, and known chemical hazards and health effects, can be combined. Because of the advanced text-mining technology, it is not only possible to search for known terms (such as known names of chemical substances), but it is also possible to use context wording (such as properties of substances) in order to identify yet unknown hazards. Results and Discussion A query typically results in a table-like overview of possible hazards, with a reference to the (scientific) sources. These overviews will then be evaluated by experts. These enhanced overviews form the input for risk management decisions. The power of this system is its ability to find any possible linguistic relationship in documents between the various terms contained in the knowledge parameters for an occupational hygiene case will be shown.