Abstract
The task of understanding customers is complex but crucial to all businesses. Identifying customer needs and wants can be simplified with the application of natural language processing methods to customer data such as customer reviews. Popular approaches to customer analysis include named entity recognition and sentiment analysis. Emotion cause extraction can be useful in analysing customer reviews to identify the reason behind a good or bad review, uncovering the needs and wants of customers. This paper explores the existing machine learning solutions available for businesses to use to understand customers better. Utilising the concept of emotion cause extraction, this paper proposes sentiment cause extraction to identify customer needs and wants from customer reviews. A dataset of customer reviews labelled with sentiment and causal spans is also introduced. The results from this paper exemplify the use of sentiment cause extraction in understanding customer needs and wants. The dataset used in this paper is publicly available at https://github.com/declare-lab/sentiment-cause-extraction