Abstract:The recent success of extractive summarization depends on the availability of large-scale annotated datasets. Existing unsupervised approaches are mostly directed graph based by combining location information with centrality computing. These methods tend to generate summaries with two problems, one is low topic coverage of the source document called the facet bias problem, and the other is continuous position distribution of extracted sentences called the position bias problem. To solve these problems, we propose the topic-aware centrality-based summarization method (TACSUM). Specifically, we employ clustering techniques to explicitly model the topics of the document and define the metrics for topic consistency and topic coverage to improve the performance of summarization. The metric topic consistency is used to guide the calculation of centrality, which solves the position bias problem and achieves a more general effect in different scenarios. We combine the metric topic coverage with the centrality to enhance the topic awareness of the model, which ensures the selected sentences are important and diverse. Numerical experimental results on four datasets show that our method outperforms previous unsupervised methods, especially in long document domains. Extensive analyses confirm that our method can generate high-quality summaries by eliminating position bias and facet bias problems.

Automatic Labeling of Topic Models Using Text Summaries

Using Topic Labels for Text Summarization.

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

Automatic Labelling Of Topic Models Learned From Twitter By Summarisation

Topic Modeling Based Text Summarization Approach

A Topic Modeling Based Approach to Novel Document Automatic Summarization

Topic Analysis for Topic-Focused Multi-Document Summarization

Topic-Aware Modeling for Unsupervised Extractive Summarization

Multi-Document Summarization Using Sentence-based Topic Models.

Chinese Multi-Document Summarization System Based on Topic Information

Bayesian Text Classification and Summarization Via A Class-Specified Topic Model.

Topic-Driven Multi-document Summarization

Automatic Labeling Of Topic Models Using Graph-Based Ranking

Subtopic-Based Multimodality Ranking for Topic-Focused Multidocument Summarization.

Subtopic-based Multi-documents Summarization

AN AUTOMATIC SUMMARIZATION APPROACH BASED ON LDA TOPIC FEATURE

Tag-Weighted Topic Model for Mining Semi-Structured Documents.

A Hybrid Topic Model for Multi-Document Summarization

Automatic Labeling Hierarchical Topics

Topic Decomposition and Summarization.

Research on Automatic Text Summarization Technology Based on ALBERT-TextRank