Train Once, Test Anywhere: Zero-Shot Learning for Text Classification
This work addresses the problem of adapting text classifiers to new classes without retraining for NLP practitioners, but it is incremental as it builds on existing zero-shot learning concepts.
The paper tackles zero-shot learning for text classification by training models to predict relationships between sentences and tag embeddings, enabling generalization to unseen classes and datasets without retraining. The models achieve competitive accuracy on test sets from the training dataset and two other standard datasets, though they do not match state-of-the-art supervised models.
Zero-shot Learners are models capable of predicting unseen classes. In this work, we propose a Zero-shot Learning approach for text categorization. Our method involves training model on a large corpus of sentences to learn the relationship between a sentence and embedding of sentence's tags. Learning such relationship makes the model generalize to unseen sentences, tags, and even new datasets provided they can be put into same embedding space. The model learns to predict whether a given sentence is related to a tag or not; unlike other classifiers that learn to classify the sentence as one of the possible classes. We propose three different neural networks for the task and report their accuracy on the test set of the dataset used for training them as well as two other standard datasets for which no retraining was done. We show that our models generalize well across new unseen classes in both cases. Although the models do not achieve the accuracy level of the state of the art supervised models, yet it evidently is a step forward towards general intelligence in natural language processing.