Convolutional Neural Network for CORE Sections Identification in Scientific Research Publications

The overwhelming volume of data generated online continuous to grow at an exponential rate. Over 80% of such data is unstructured. Scientific research publications constitute a significant portion of such unstructured data. Systematic literature review (SLR) is a rigorous and challenging process. The key challenge in SLR activity is the automatic extraction of the relevant data from the volume of publications. Lack of a unified framework has been identified as the key problem. A canonical model, based on the structure of the papers was proposed for data extraction purposes in SLR. Implemented as a classification problem, traditional machine learning models were used to realize the canonical model. A good accuracy was reported in these traditional models. However, there is room for improvement. This paper presents the result of the investigation of the same problem using convolutional neural network (CNN), which is more sophisticated (deeper). The results show an improvement over the traditional machine learning models with an accuracy of 85%. Unlike the previous CNN NLP works, this work also demonstrates the application of CNN on a bigger NLP dataset such as the data from the scientific research publications. The result also shows that the CNN performs even better in NLP tasks with bigger datasets.