Octa: Omissions and Conflicts in Target-Aspect Sentiment Analysis
This addresses sentiment analysis for opinionated text by improving accuracy in handling target-aspect conflicts, though it is incremental as it builds on existing attention mechanisms.
The paper tackles the problem of conflicting sentiments between targets and aspects in opinionated text by proposing Octa, a model that jointly considers both when inferring sentiments, resulting in accuracy gains of 1.6% to 4.3% over leading models on benchmark datasets.
Sentiments in opinionated text are often determined by both aspects and target words (or targets). We observe that targets and aspects interrelate in subtle ways, often yielding conflicting sentiments. Thus, a naive aggregation of sentiments from aspects and targets treated separately, as in existing sentiment analysis models, impairs performance. We propose Octa, an approach that jointly considers aspects and targets when inferring sentiments. To capture and quantify relationships between targets and context words, Octa uses a selective self-attention mechanism that handles implicit or missing targets. Specifically, Octa involves two layers of attention mechanisms for, respectively, selective attention between targets and context words and attention over words based on aspects. On benchmark datasets, Octa outperforms leading models by a large margin, yielding (absolute) gains in accuracy of 1.6% to 4.3%.