SemEval 2014 is a collection of datasets used for the Semantic Evaluation (SemEval) workshop, an annual event that focuses on the evaluation and comparison of systems that can analyze diverse semantic phenomena in text. The datasets from SemEval 2014 are used for various tasks, including but not limited to:
- Aspect-Based Sentiment Analysis (ABSA): This task is based on laptop and restaurant reviews. It involves identifying the aspects or features mentioned in a review and determining the sentiment expressed towards each aspect.
- Text Classification: This task involves classifying text into predefined categories. Sub-tasks include text-scoring, natural language inference, and semantic-similarity-scoring.