Abstract
With the advent of the internet and social media, large volumes of customer experience data is generated on a day-to-day basis. Understanding and extracting key insights from this data is critical for enterprises and organizations to manage and improve the experience for their customers.
The aim of this project was to apply machine learning (ML) and natural language processing (NLP) in order to extract the key entities (referred to as aspects) and the sentiment expressed towards them to better understand the emotion of the customer.
The novelty of the project came from developing innovative state-of-the-art methods to accurately identify the entities, contexts related to the entities and also the sentiment about them. In particular, we explored how ML models can be influenced by background knowledge extracted from dependency parsers and generative lexicons. Our recent work suggested that effective ways to discover the context in which a sentiment is being expressed towards the aspects is crucial to improving the accuracy of sentiment analysis.
Research Staff
- Nirmalie Wiratunga
- Stewart Massie
- Anil Bandhakavi
- Rushi Luhar
- Sharad Khandelwal