Enhancing the Interoperability of iSimp by Using the BioC Format

*Corresponding author: Tel: 302 831 8496, E-mail: [email protected] ! Abstract This paper reports the use of the BioC format in our sentence simplification system, iSimp, so that it could be seamlessly used in text mining pipelines. iSimp is designed to simplify complex sentences commonly found in the biomedical text, therefore bringing benefits to existing text mining applications that rely on the analysis of sentence structures. By adopting the BioC format, we aim to make iSimp readily integrable in various applications in this domain. To examine the utility of iSimp with BioC, we designed and implemented a rule-based relation extraction system that uses iSimp as a preprocessing module and BioC format for data exchange. Evaluation on the BioNLP-ST 2011 GE task training corpus showed that, with sentence simplification provided by iSimp, the F-value of the phosphorylation extraction increased 3%. The iSimp corpus previously used for the evaluation of simplification and the GE task corpus used in the current study have been converted into the BioC format and made publicly available. ! Introduction The syntactic complexity of the biomedical text often poses a major challenge in designing and applying Natural Language Processing (NLP) systems on scientific articles. One possible approach to address this issue and improve the performance of NLP systems (e.g., relation extraction systems) is to simplify the complexity of the sentences themselves prior to using them as input in the NLP systems. For this purpose, we had previously developed iSimp [1], a sentence simplification system. Used a preprocessing module that simplifies the input text, iSimp has a potential to enhance existing text mining applications in the biomedical domain. In order to make iSimp readily integrable in various applications, we have adopted the BioC format, a simple, yet robust, XML format to share text documents and annotations [2].