Parameter extraction from technical datasheet documentation using technique optimized through genetic algorithm

The paper presents an approach for extracting specific parameters from technical datasheet documentation. The extraction is aimed at various electrical properties of different electronic components. The presented approach is part of an automated data gathering system that is used for the composition of a complex database of electronic components. Key variables of the proposed approach are optimized through the implementation a genetic algorithm. The paper presents the specifics and challenges related to the extraction, the steps taken to resolve them and optimize accuracy, as well as verification and assessment applied over 8000 documents.