Target-Based Sentiment Analysis as a Sequence-Tagging Task
This research attempts to provide a complete solution to the sentiment analysis task (extraction of opinion holder, polarity of the underlying sentiment and target) by treating it as a sequence tagging problem.The best performing models are LSTM-based. We also address class imbalance and the need for meaningful data augmentation techniques to increase the size of the training set and make the use of LSTMs possible.
Gerolemou has recently graduated with an MSc. in Artificial Intelligence from Maastricht University. Additionally, she holds a BSc. in Computing Science from the University of Glasgow. She works as a Data Scientist/Software Developer at ZyLAB, Amsterdam.