پریناز سلطان زاده

Relation extraction between drugs plays a crucial role in identifying drug–drug interactions and predicting side effects. The advancement of machine learning methods in relation extraction, along with the development of large medical text databases, has enabled the low-cost extraction of such relations compared to other approaches that typically require expert knowledge. However, to the best of our knowledge, there are limited datasets specifically designed for drug–drug relation extraction currently available. Therefore, employing transfer learning becomes necessary to apply machine learning methods in this domain. In this study, we propose DREAM, a method that first employs a trained relation extraction model to discover relations between entities and then applies this model to a corpus of medical texts to construct an ontology of drug relationships. The main contribution of this study is adapting the ACORD model to the medical domain using transfer learning, enabling the extraction of domain-specific drug–drug relations and the construction of an ontology. The extracted relations are subsequently validated using a large language model. Quantitative results indicate that the LLM agreed with 71 % of the relations extracted from a subset of PubMed abstracts. Furthermore, our qualitative analysis indicates that this approach can uncover ambiguities in the medical domain, highlighting the challenges inherent in relation extraction in this field.

Multi-label text classification is a critical challenge in natural language processing, where the goal is to assign multiple labels to a given document. Recent advances have primarily focused on deep learning approaches, yet many fail to adequately capture the intricate relationships between documents and labels. In this paper, we propose a novel method called MultiCGCN, in which we leverage Graph Convolutional Networks (GCNs) for multi-label text classification by modeling text as a heterogeneous graph. This unified graph incorporates document similarities, label relationships, and document-label associations, enabling the model to effectively capture both document and label dependencies. We transform the multi-label classification problem into a link prediction task, using Term Frequency–Inverse Document Frequency (TF-IDF) for document similarity and applying GCNs to predict label assignments. Our empirical evaluations demonstrate that MultiCGCN achieves a significant performance boost, improving F1 score by 10% over traditional baseline models. This approach opens new avenues for enhancing the accuracy of multi-label classification in various domains.