We explore the notion of uncertainty in the context of modern abstractive summarization models, using the tools of Bayesian Deep Learning. Our approach approximates Bayesian inference by first extending state-of-the-art summarization models with Monte Carlo dropout and then using them to perform multiple stochastic forward passes. Based on Bayesian inference we are able to effectively quantify uncertainty at prediction time. Having a reliable uncertainty measure, we can improve the experience of the end user by filtering out generated summaries of high uncertainty. Furthermore, uncertainty estimation could be used as a criterion for selecting samples for annotation, and can be paired nicely with active learning and human-in-the-loop approaches. Finally, Bayesian inference enables us to find a Bayesian summary which performs better than a deterministic one and is more robust to uncertainty. In practice, we show that our Variational Bayesian equivalents of BART and PEGASUS can outperform their deterministic counterparts on multiple benchmark datasets.

Recent Deep Learning (DL) summarization models greatly outperform traditional summarization methodologies, generating high-quality summaries. Despite their success, there are still important open issues, such as the limited engagement and trust of users in the whole process. In order to overcome these issues, we reconsider the task of summarization from a human-centered perspective. We propose to integrate a user interface with an underlying DL model, instead of tackling summarization as an isolated task from the end user. We present a novel system, where the user can actively participate in the whole summarization process. We also enable the user to gather insights into the causative factors that drive the model’s behavior, exploiting the self-attention mechanism. We focus on the financial domain, in order to demonstrate the efficiency of generic DL models for domain-specific applications. Our work takes a first step towards a model-interface co-design approach, where DL models evolve along user needs, paving the way towards human-computer text summarization interfaces.

We present the systems we submitted for the shared tasks of the Workshop on Scholarly Document Processing at EMNLP 2020. Our approaches to the tasks are focused on exploiting large Transformer models pre-trained on huge corpora and adapting them to the different shared tasks. For tasks 1A and 1B of CL-SciSumm we are using different variants of the BERT model to tackle the tasks of “cited text span” and “facet” identification. For the summarization tasks 2 of CL-SciSumm, LaySumm and LongSumm we make use of different variants of the PEGASUS model, with and without fine-tuning, adapted to the nuances of each one of those particular tasks.