Attention Models
Related Papers
Textual Entailment with Structured Attentions and Composition
AbstractDeep learning techniques are increasingly popular in the textual entailment task, overcoming the fragility of traditional discrete models with hard alignments and logics. In particular, the recently proposed attention models …...
A Position-aware Bidirectional Attention Network for Aspect-level Sentiment Analysis
AbstractAspect-level sentiment analysis aims to distinguish the sentiment polarity of each specific aspect term in a given sentence. Both industry and academia have realized the importance of the relationship between …...
On the Cost-Effectiveness of Stacking of Neural and Non-Neural Methods for Text Classification: Scenarios and Performance Prediction
Neural network algorithms such as those based on transformers and attention models have ex- celled on Automatic Text Classification (ATC) tasks. However, such enhanced performance comes at high computational costs. …...
UTNLP at SemEval-2022 Task 6: A Comparative Analysis of Sarcasm Detection Using Generative-based and Mutation-based Data Augmentation
AbstractSarcasm is a term that refers to the use of words to mock, irritate, or amuse someone. It is commonly used on social media. The metaphorical and creative nature of …...