No abstract

Procedural text, which describes entities and their interactions as they undergo some process, depicts entities in a uniquely nuanced way. First, each entity may have some observable discrete attributes, such as its state or location; modeling these involves imposing global structure and enforcing consistency. Second, an entity may have properties which are not made explicit but can be effectively induced and tracked by neural networks. In this paper, we propose a structured neural architecture that reflects this dual nature of entity evolution. The model tracks each entity recurrently, updating its hidden continuous representation at each step to contain relevant state information. The global discrete state structure is explicitly modelled with a neural CRF over the changing hidden representation of the entity. This CRF can explicitly capture constraints on entity states over time, enforcing that, for example, an entity cannot move to a location after it is destroyed. We evaluate the performance of our proposed model on QA tasks over process paragraphs in the PROPARA dataset and find that our model achieves state-of-the-art results.

Tracking entities in procedural language requires understanding the transformations arising from actions on entities as well as those entities' interactions. While self-attention-based pre-trained language encoders like GPT and BERT have been successfully applied across a range of natural language understanding tasks, their ability to handle the nuances of procedural texts is still untested. In this paper, we explore the use of pre-trained transformer networks for entity tracking tasks in procedural text. First, we test standard lightweight approaches for prediction with pre-trained transformers, and find that these approaches underperform even simple baselines. We show that much stronger results can be attained by restructuring the input to guide the transformer model to focus on a particular entity. Second, we assess the degree to which transformer networks capture the process dynamics, investigating such factors as merged entities and oblique entity references. On two different tasks, ingredient detection in recipes and QA over scientific processes, we achieve state-ofthe-art results, but our models still largely attend to shallow context clues and do not form complex representations of intermediate entity or process state. 1

The recent outbreak of novel coronavirus disease (COVID-19) has resulted in healthcare crises across the globe. Moreover, the persistent and prolonged complications of post-COVID-19 or long COVID are also putting extreme pressure on hospital authorities due to the constrained healthcare resources. Out of many long-lasting post-COVID-19 complications, heart disease has been realized as the most common among COVID-19 survivors. The motivation behind this research is the limited availability of the post-COVID-19 dataset. In the current research, data related to post-COVID complications are collected by personally contacting the previously infected COVID-19 patients. The dataset is preprocessed to deal with missing values followed by oversampling to generate numerous instances, and model training. A binary classifier based on a stacking ensemble is modeled with deep neural networks for the prediction of heart diseases, post-COVID-19 infection. The proposed model is validated against other baseline techniques, such as decision trees, random forest, support vector machines, and artificial neural networks. Results show that the proposed technique outperforms other baseline techniques and achieves the highest accuracy of 93.23%. Moreover, the results of specificity (95.74%), precision (95.24%), and recall (92.05%) also prove the utility of the adopted approach in comparison to other techniques for the prediction of heart diseases.

No abstract