Past Event! Note: this event has already taken place.
Aspect-based Sentiment Analysis and its Applications
February 14, 2019 at 10:00 AM to 11:00 AM
|Location:||5345 Herzberg Laboratories|
Aspect-based sentiment analysis (ABSA) has recently attracted increasing research attention. Most previous work relies on syntax, such as automatic parse trees, which are subject to noise for informal text such as tweets and social media. In this project, we show that competitive results can be achieved without the use of syntax, by extracting a rich set of automatic features relied on a pattern-based approach. In particular, we split an input text into a left context and a right context according to a given target to extract features. In standard evaluation, the conceptually simple method gives a 4.8% absolute improvement over the state-of-the-art on three-way targeted sentiment classification, achieving the best reported results for this task. Most of the current works follow this approach to extract its representation and show that this is the most suitable method to solve the problem. In addition, we also present an application of the ABSA in the hotel domain by building an end-to-end system. The system is able to simultaneously stream reviews online and predict sentiment towards specific aspects. Finally, a summarized report is automatically obtained to provide to clients.
Tin D. Vo is currently a data scientist at Chata.ai. Before going to Laval University to work as a postdoctoral researcher, he obtained the PhD degree in Computer Science under the supervision of Professor Yue Zhang at Singapore University of Technology and Design (SUTD) in September 2017. He had two-year experience as a lecturer at Can Tho University after getting his Bachelor degree there in 2011. His research interests include natural language processing (NLP), machine learning and artificial intelligence. He has been focusing on applying machine learning and deep learning into sentiment analysis, E-community toxicity and text classification. He is recognized in the field by having several publications on the top conferences in NLP (i.e. ACL, EMNLP, CICLing) and AI (i.e. AAAI, IJCAI). After graduating, he has been centering on employing what he was studied into production by constructing the most efficient and accurate text classifiers which are currently used in industry such as language detection and aspect-based sentiment analysis. In addition, he is focusing on developing cutting-edge neural network models in NLP such as text classification, sentiment analysis, sequence tagging, and topic modelling.
No registration required, seating is on a first come, first serve basis.