Personal Research Topics
Data Interoperability Services in Cloud Infrastructures and Policy Making Applications
Approach primarily focuses on the phases of the translation, the processing, the annotation, the mapping, as well as the transformation of the collected data, which have major impact on the successful aggregation, analysis, and exploitation of da-ta across the whole policy making lifecycle
Neural Machine Translation (NMT) with an emphasis on Transfer and Federated Learning methods
The application of NMT techniques enhanced by the use of Transfer Learning will allow the transfer of exported results, parameters, and knowledge from one NMT model to another, with the ultimate goal of developing and optimizing multilingual classifiers. Furthermore, utilizing Federated Learning techniques will improve, enhance, and further extend the capabilities and interoperability of the proposed NMT system.
Multilingual Sentiment Analysis
Despite the evolution and the advancements in neural sentiment analysis most approaches for sentence-level neural sentiment analysis available today were developed only for English and there are only a few efforts that approach the problem considering other and especially low-interest and usage languages. To this end, many approaches, and researches over the past five years enhance multilingual sentiment analysis based on encoder-decoder model and the utilization of NMT techniques.
