Alert button

Aspect-Controllable Opinion Summarization

Sep 07, 2021
Reinald Kim Amplayo, Stefanos Angelidis, Mirella Lapata

Figure 1 for Aspect-Controllable Opinion Summarization
Figure 2 for Aspect-Controllable Opinion Summarization
Figure 3 for Aspect-Controllable Opinion Summarization
Figure 4 for Aspect-Controllable Opinion Summarization

Share this with someone who'll enjoy it:

Recent work on opinion summarization produces general summaries based on a set of input reviews and the popularity of opinions expressed in them. In this paper, we propose an approach that allows the generation of customized summaries based on aspect queries (e.g., describing the location and room of a hotel). Using a review corpus, we create a synthetic training dataset of (review, summary) pairs enriched with aspect controllers which are induced by a multi-instance learning model that predicts the aspects of a document at different levels of granularity. We fine-tune a pretrained model using our synthetic dataset and generate aspect-specific summaries by modifying the aspect controllers. Experiments on two benchmarks show that our model outperforms the previous state of the art and generates personalized summaries by controlling the number of aspects discussed in them.

* EMNLP 2021  
View paper onarxiv icon

Share this with someone who'll enjoy it: