Topic Segmentation in the Wild: Towards Segmentation of Semi-structured & Unstructured Chats
This addresses the challenge of segmenting unstructured conversations for NLP applications, but it is incremental as it focuses on generalization analysis.
The paper tackled the problem of topic segmentation for unstructured texts, finding that pre-training on structured text does not help, while training from scratch on a small unstructured dataset significantly improves results.
Breaking down a document or a conversation into multiple contiguous segments based on its semantic structure is an important and challenging problem in NLP, which can assist many downstream tasks. However, current works on topic segmentation often focus on segmentation of structured texts. In this paper, we comprehensively analyze the generalization capabilities of state-of-the-art topic segmentation models on unstructured texts. We find that: (a) Current strategies of pre-training on a large corpus of structured text such as Wiki-727K do not help in transferability to unstructured texts. (b) Training from scratch with only a relatively small-sized dataset of the target unstructured domain improves the segmentation results by a significant margin.