Transformer-based Models for Long-Form Document Matching: Challenges and Empirical Analysis
Preprint
Abstract
Recent advances in the area of long document matching have primarily focused on using transformer-based models for long document encoding and matching. There are two primary challenges associated with these models. Firstly, the performance gain provided by transformer-based models comes at a steep cost - both in terms of the required training time and the resource (memory and energy) consumption. The second major limitation is their inability to handle more than a pre-defined input token length at a time. In this work, we empirically demonstrate the effectiveness of simple neural models (such as feed-forward networks, and CNNs) and simple embeddings (like GloVe, and Paragraph Vector) over transformer-based models on the task of document matching. We show that simple models outperform the more complex BERT-based models while taking significantly less training time, energy, and memory. The simple models are also more robust to variations in document length and text perturbations.Keywords
All Related Versions
This publication has 4 references indexed in Scilit:
- Supervised Contrastive Learning for Interpretable Long-Form Document MatchingACM Transactions on Knowledge Discovery From Data, 2023
- Hierarchical Transformers for Multi-Document Summarization2019
- Hierarchical Attention Networks for Document Classification2016
- The ACL anthology network corpusLanguage Resources and Evaluation, 2013