Abstract:Text summarizing is the task of reducing a document's length while maintaining its essential information. In the age of information explosion, how to obtain the content that users needed from a large volume of information becomes particularly significant. Under such circumstances, query-focused abstractive summarization (qfs) becomes more dominant since it is able to focus on user needs while delivering fluent, concise, succinct paraphrased summaries. However, unlike generic summarization, which has achieved remarkable progress driven by a substantial amount of parallel data, the qfs struggles due to a deficiency of parallel corpus. Therefore, in this paper, we leverage a typical large generic summarization dataset to facilitate the pressing demands on unsupervised qfs. The large-scale query-free benchmark is automatically transformed into a query-focused dataset (Query-CNNDM) while preserving its informative summaries. We propose a simple yet effective unsupervised method, called Domain Adaptation and Summary Distillation method (DASD). In the model, to achieve the domain adaptation for unsupervised qfs, we design a query-aware gap sentence generation (q-GSG) strategy to equip the model with the capability of learning target textual knowledge and obtaining a good initialization at the target domain. As instance-specific regularization, we train a teacher model with the Query-CNNDM to generate pseudo-labels for summary distillation. Experimental results indicate that our DASD model achieves state-of-the-art performance on two benchmark datasets, Debatepedia and Wikiref, in a zero-shot setting and shows good generalization to the abstractive few-shot qfs.

Query-focused summarization with the context-graph information fusion transformer

Query-focused Multi-document Summarization: Combining a Novel Topic Model with Graph-based Semi-supervised Learning

Exploring hypergraph-based semi-supervised ranking for query-oriented summarization

Exploring Neural Models for Query-Focused Summarization

QontSum: On Contrasting Salient Content for Query-focused Summarization

Query-Focused Multi-document Summarization Based on Concept Importance

Query-Focused Summarization by Combining Topic Model and Affinity Propagation

Nonfactoid Question Answering as Query-Focused Summarization With Graph-Enhanced Multihop Inference

A Lightweight Constrained Generation Alternative for Query-focused Summarization

Query-Utterance Attention with Joint modeling for Query-Focused Meeting Summarization

A Graph Based Query Focused Multi-Document Summarization

QuerySum: A Multi-Document Query-Focused Summarization Dataset Augmented with Similar Query Clusters

Query-Oriented Micro-Video Summarization

Abstract Summarization Model Based on Semantic Graphs and Entity Pointers

Domain Adaptation and Summary Distillation for Unsupervised Query Focused Summarization

Your Interest, Your Summaries: Query-Focused Long Video Summarization

HyperSum: hypergraph based semi-supervised sentence ranking for query-oriented summarization.

Does This Summary Answer My Question? Modeling Query-Focused Summary Readers with Rational Speech Acts

IDEAL: Leveraging Infinite and Dynamic Characterizations of Large Language Models for Query-focused Summarization

Learning to Rank Salient Content for Query-focused Summarization

Query-oriented unsupervised multi-document summarization via deep learning model