Opinion Extraction as A Structured Sentiment Analysis using Transformers
This work addresses sentiment analysis for applications requiring detailed opinion extraction, but it appears incremental as it builds on existing tasks without claiming major breakthroughs.
The paper tackled the problem of structured sentiment analysis by combining relationship extraction and named entity recognition into a single stacked model to extract opinion tuples (holders, targets, expressions) from sentences, achieving results through experiments to find the best model.
Relationship extraction and named entity recognition have always been considered as two distinct tasks that require different input data, labels, and models. However, both are essential for structured sentiment analysis. We believe that both tasks can be combined into a single stacked model with the same input data. We performed different experiments to find the best model to extract multiple opinion tuples from a single sentence. The opinion tuples will consist of holders, targets, and expressions. With the opinion tuples, we will be able to extract the relationship we need.