SSAT: Scientific Storyboarding Framework Using Artificial Intelligence Techniques

This paper introduces a scientific storyboarding framework that uses artificial intelligence for learning and inferencing. It works by classifying a dataset of documents using the Bi-LSTM classifier to identify different instances. The results are then used to create metadata, which is further classified using deep belief networks. Next, the user's topic of interest is pre-processed, and an intermediate ontology model is generated. This model helps build knowledge by gathering information from the World Wide Web using recurrent neural networks. The framework then combines the results from two distinct classifiers—one from the dataset and the other from the user's query. These results are analyzed through explicit semantic analysis to select relevant entities. The monkey algorithm is utilized for optimization, resulting in an optimal solution set. To ensure semantic reasoning and entity regulation, Morisita index and explicit semantic analysis is applied. The proposed SSAT model achieves high percentages of precision, recall, accuracy, and F-measure, specifically 95.58, 96.83, 96.205, and 96.2009397%, respectively. Additionally, the model achieves a low value of FDR (False Discovery Rate) at 0.05%.