# Self-Adaptive Hierarchical Sentence Model

arXiv: https://arxiv.org/abs/1504.05070 · HTML: https://mlsanity.com/paper/1504.05070/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Apr 2015.

## Built on by (most cited first, 10 of 10)

- [Revisiting Uncertainty-based Query Strategies for Active Learning with Transformers](https://mlsanity.com/paper/2107.05687/) (2021)
- [Local Structure Matters Most: Perturbation Study in NLU](https://mlsanity.com/paper/2107.13955/) (2021)
- [Czech Dataset for Cross-lingual Subjectivity Classification](https://mlsanity.com/paper/2204.13915/) (2022)
- [A Survey on Semantic Processing Techniques](https://mlsanity.com/paper/2310.18345/) (2023)
- [Deep Learning-based Sentiment Classification: A Comparative Survey](https://mlsanity.com/paper/2312.17253/) (2023)
- [Polarity and Subjectivity Detection with Multitask Learning and BERT Embedding](https://mlsanity.com/paper/2201.05363/) (2022)
- [Semantic Answer Type and Relation Prediction Task (SMART 2021)](https://mlsanity.com/paper/2112.07606/) (2021)
- [Integrating Dependency Tree Into Self-attention for Sentence Representation](https://mlsanity.com/paper/2203.05918/) (2022)
- [Embedding Compression for Text Classification Using Dictionary Screening](https://mlsanity.com/paper/2211.12715/) (2022)
- [Analyzing the Generalizability of Deep Contextualized Language Representations For Text Classification](https://mlsanity.com/paper/2303.12936/) (2023)

## Abstract (arXiv)

The ability to accurately model a sentence at varying stages (e.g., word-phrase-sentence) plays a central role in natural language processing. As an effort towards this goal we propose a self-adaptive hierarchical sentence model (AdaSent). AdaSent effectively forms a hierarchy of representations from words to phrases and then to sentences through recursive gated local composition of adjacent segments. We design a competitive mechanism (through gating networks) to allow the representations of the same sentence to be engaged in a particular learning task (e.g., classification), therefore effectively mitigating the gradient vanishing problem persistent in other recursive models. Both qualitative and quantitative analysis shows that AdaSent can automatically form and select the representations suitable for the task at hand during training, yielding superior classification performance over competitor models on 5 benchmark data sets.