SECTOR: A Neural Model for Coherent Topic Segmentation and Classification
arXiv:1902.04793
Abstract
When searching for information, a human reader first glances over a document, spots relevant sections and then focuses on a few sentences for resolving her intention. However, the high variance of document structure complicates to identify the salient topic of a given section at a glance. To tackle this challenge, we present SECTOR, a model to support machine reading systems by segmenting documents into coherent sections and assigning topic labels to each section. Our deep neural network architecture learns a latent topic embedding over the course of a document. This can be leveraged to classify local topics from plain text and segment a document at topic shifts. In addition, we contribute WikiSection, a publicly available dataset with 242k labeled sections in English and German from two distinct domains: diseases and cities. From our extensive evaluation of 20 architectures, we report a highest score of 71.6% F1 for the segmentation and classification of 30 topics from the English city domain, scored by our SECTOR LSTM model with bloom filter embeddings and bidirectional segmentation. This is a significant improvement of 29.5 points F1 compared to state-of-the-art CNN classifiers with baseline segmentation.
Author's final version, accepted for publication at TACL, 2019
References in corpus (5)
- Distributed Representations of Sentences and Documents
- Finding community structure in networks using the eigenvectors of matrices
- TopicRNN: A Recurrent Neural Network with Long-Range Semantic Dependency
- Text Segmentation based on Semantic Word Embeddings
- Automatic Labelling of Topics with Neural Embeddings