Argument mining is the natural language processing (NLP) technique that aims to automatically identify and extract the structures of interpretation and reasoning behind the controversial conservation. It is important to consider the contextual information in argument mining to understand better the topic discussed in the conversation. Context-aware argument mining provides the potential information to identify the relationships between the conversation.

  Traditional supervised learning models for argument mining are error-prone and domain-dependent. Semi-supervised learning models are independent of the domain and provide better identification of arguments. Deep semi-supervised learning effectively utilizes labeled and unlabeled data to make predictions using a deep neural network in large-scale datasets. Deep semi-supervised learning-based context-aware argument mining is robust that perfectly grasps the potential argument and improves the performance of the argumentation mining system.

  • In argumentation mining, context is signified as the decisive criteria for identifying arguments and argumentative relations in textual information classification.

  • Argumentation mining is predicated on features extracted from context segments, upgrading state-of-the-art argument components to enhance significantly.

  • The contextual information sources are shallow word embeddings, knowledge graphs, and fine-tuned transfer learning.

  • Annotation of the arguments is highly context-dependent and error-prone.

  • Semi-supervised learning approach utilizes clustering of unlabeled data with innovative features to boost the performance of argument component identification in argument mining.