iSimp in BioC standard format: enhancing the interoperability of a sentence simplification system

Area of Science:

• Biomedical Natural Language Processing
• Computational Biology
• Bioinformatics

Background:

• Biomedical texts contain complex sentences that hinder text mining.
• Existing text mining applications struggle with sentence structure analysis.
• Interoperability between biomedical tools is crucial for efficient data exchange.

Purpose of the Study:

• To demonstrate the utility and generalizability of the iSimp sentence simplification system.
• To integrate iSimp with the BioC standard format for enhanced interoperability.
• To evaluate the impact of iSimp sentence simplification on a relation extraction task.

Main Methods:

• The iSimp system was employed as a preprocessing module within a rule-based relation extraction system.
• The BioC format was utilized for data exchange and interoperability.
• A rule-based relation extraction system was implemented and evaluated on the BioNLP-ST 2011 GENIA Event Extraction corpus.

Main Results:

• iSimp sentence simplification improved recall by 3.2% in the GENIA Event Extraction task.
• Precision was maintained without reduction during the evaluation.
• The iSimp-annotated corpora were converted to BioC format, enhancing accessibility.

Conclusions:

• The iSimp system effectively simplifies complex biomedical sentences.
• Adoption of the BioC format enhances the interoperability of iSimp with other biomedical applications.
• iSimp shows significant utility in improving text mining tasks, particularly relation extraction.