Evaluation toolkit for ontology-based event extraction systems

• Main Author: Oo, Alvin Ye Kei.
• Other Authors: School of Computer Engineering
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/52021
• Institution: Nanyang Technological University

When text mining becomes more popular,we need a kind of software which developer can retrieve and research details so that they can kepep within their own fields of researches.BioNLP shared task is which promotes and extract events from unstructured text,so that the participants will be able to retrieve the results for evaluations purpose. Brat annotation toolkit was recently introduced,although it has an inituitive and user-friedndly interfaces,one of our goals is to make annotation more accessible to the non-technical users,thus it will reduce and maintain the human and financial cost of annotations,along the way,it helps to make the development of the job more accessible to the user. The project itself is not trival,considering the fact that that other’s has not so many concepts. Using brat annotation software allows the annotator to link up concepts and events together,and also display the violation.However,there is a problem,different annotator may missed out several different concepts,hence we plan to solve the problem because two different annotators may not have agreed on the same thing.Precision,recall,and F-scores is used as a performance measure for the systems.In additional,we have also tested a simple event trigger recongition method.