The analysis of social discourses from the perspective of historical changes deserves special attention. Such a study could play a key role in revealing social changes and latent narrative of those in power; and understanding the underlying social dynamic in a given period. Until the recent years, such issues were analyzed mainly in a qualitative approach. In our paper we present a new way of revealing/discovering and interpreting social discourses using an advanced NLP method called word embedding. Based on word similarities we can understand the main structural frames of a given system and using a dynamic approach we can reveal the social changes in a historical period. In our study we created a large corpus from the Hungarian "Pártélet" journal (1956–89). This was the official journal of the governing party, hence it represents not just a media discourse of the era, but the official discourse of the government, too. One of the main focal points of our research is to study the evolution of the semantic content of some of the concepts related to the topics of agriculture and industry, which are two central notions of the examined era.
The analysis of social discourses from the perspective of historical changes deserves special attention. Such a study could play a key role in revealing social changes and latent narrative of those in power; and understanding the underlying social dynamic in a given period. Until the recent years, such issues were analyzed mainly in a qualitative approach. In our paper we present a new way of revealing/discovering and interpreting social discourses using an advanced NLP method called word embedding. Based on word similarities we can understand the main structural frames of a given system and using a dynamic approach we can reveal the social changes in a historical period. In our study we created a large corpus from the Hungarian "Pártélet" journal (1956–89). This was the official journal of the governing party, hence it represents not just a media discourse of the era, but the official discourse of the government, too. One of the main focal points of our research is to study the evolution of the semantic content of some of the concepts related to the topics of agriculture and industry, which are two central notions of the examined era.
Developing artificial learning systems that can understand and generate natural language has been one of the long-standing goals of artificial intelligence. Recent decades have witnessed an impressive progress on both of these problems, giving rise to a new family of approaches. Especially, the advances in deep learning over the past couple of years have led to neural approaches to natural language generation (NLG). These methods combine generative language learning techniques with neural-networks based frameworks. With a wide range of applications in natural language processing, neural NLG (NNLG) is a new and fast growing field of research. In this state-of-the-art report, we investigate the recent developments and applications of NNLG in its full extent from a multidimensional view, covering critical perspectives such as multimodality, multilinguality, controllability and learning strategies. We summarize the fundamental building blocks of NNLG approaches from these aspects and provide detailed reviews of commonly used preprocessing steps and basic neural architectures. This report also focuses on the seminal applications of these NNLG models such as machine translation, description generation, automatic speech recognition, abstractive summarization, text simplification, question answering and generation, and dialogue generation. Finally, we conclude with a thorough discussion of the described frameworks by pointing out some open research directions. ; This work has been partially supported by the European Commission ICT COST Action "Multi-task, Multilingual, Multi-modal Language Generation" (CA18231). AE was supported by BAGEP 2021 Award of the Science Academy. EE was supported in part by TUBA GEBIP 2018 Award. BP is in in part funded by Independent Research Fund Denmark (DFF) grant 9063-00077B. IC has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 838188. EL is partly funded by Generalitat Valenciana and the Spanish Government throught projects PROMETEU/2018/089 and RTI2018-094649-B-I00, respectively. SMI is partly funded by UNIRI project uniri-drustv-18-20. GB is partly supported by the Ministry of Innovation and the National Research, Development and Innovation Office within the framework of the Hungarian Artificial Intelligence National Laboratory Programme. COT is partially funded by the Romanian Ministry of European Investments and Projects through the Competitiveness Operational Program (POC) project "HOLOTRAIN" (grant no. 29/221 ap2/07.04.2020, SMIS code: 129077) and by the German Academic Exchange Service (DAAD) through the project "AWAKEN: content-Aware and netWork-Aware faKE News mitigation" (grant no. 91809005). ESA is partially funded by the German Academic Exchange Service (DAAD) through the project "Deep-Learning Anomaly Detection for Human and Automated Users Behavior" (grant no. 91809358).
