Dear Colleagues,

Recent advances in natural language processing (NLP) that involve machine learning and deep learning have certainly revolutionized the field. Still, there are specific tasks and domains where these new techniques have still not surpassed more classical approaches—for example, tasks that require deep linguistic knowledge such as natural language understanding, semantic reasoning, and question answering. Another common limitation is that of the scarcity of training datasets, a situation that arises when trying to apply recent approaches to new domains. To overcome these limitations, it is necessary to consider hybrid systems that exploit domain-oriented knowledge into learning models in a way that allows machines to grasp the intricacies of real-world applications, equipping them with deep understanding and general common sense.

While there are efforts to design hybrid models, several aspects need to be considered. such as interpretability, transparency, accountability, and efficiency. This Special Issue of Electronics addresses the direction of NLP efforts toward hybrid solutions, considering the mentioned characteristics and their effects on end users and society in general.

Topics of interest of this Special Issue include but are not limited to:

• Information extraction;
• Semantic reasoning;
• Text and speech processing;
• Relational semantics;
• Discourse analysis;
• Argument mining;
• Text summarization;
• Machine translation;
• Natural language generation;
• Natural language understanding;
• Question answering;
• Sentiment and emotion analysis;
• Affect analysis;
• Hate speech analysis;
• Radicalization analysis;
• Disinformation analysis;
• Authorship attribution.

Dr. Oscar Araque
Dr. Lorenzo Gatti
Dr. Álvaro Carrera Barroso
Dr. Kyriaki Kalimeri
Guest Editors