Text Length Adaptation in Sentiment Classification
Preprint
Abstract
Can a text classifier generalize well for datasets where the text length is different? For example, when short reviews are sentiment-labeled, can these transfer to predict the sentiment of long reviews (i.e., short to long transfer), or vice versa? While unsupervised transfer learning has been well-studied for cross domain/lingual transfer tasks, Cross Length Transfer (CLT) has not yet been explored. One reason is the assumption that length difference is trivially transferable in classification. We show that it is not, because short/long texts differ in context richness and word intensity. We devise new benchmark datasets from diverse domains and languages, and show that existing models from similar tasks cannot deal with the unique challenge of transferring across text lengths. We introduce a strong baseline model called BaggedCNN that treats long texts as bags containing short texts. We propose a state-of-the-art CLT model called Length Transfer Networks (LeTraNets) that introduces a two-way encoding scheme for short and long texts using multiple training mechanisms. We test our models and find that existing models perform worse than the BaggedCNN baseline, while LeTraNets outperforms all models.Keywords
All Related Versions
This publication has 23 references indexed in Scilit:
- Adversarial Deep Averaging Networks for Cross-Lingual Sentiment ClassificationTransactions of the Association for Computational Linguistics, 2018
- Multiple Instance Learning Networks for Fine-Grained Sentiment AnalysisTransactions of the Association for Computational Linguistics, 2018
- Aspect Sentiment Model for Micro Reviews2017
- Learning Sentence Embeddings with Auxiliary Tasks for Cross-Domain Sentiment Classification2016
- From Group to Individual Labels Using Deep Features2015
- Document Modeling with Gated Recurrent Neural Network for Sentiment Classification2015
- Transfer Learning for Text Mining2012
- Cross-domain sentiment classification via spectral feature alignment2010
- Multi-instance learning by treating instances as non-I.I.D. samples2009
- Opinion Mining and Sentiment AnalysisFoundations and Trends® in Information Retrieval, 2008