Unsupervised Extractive Summarization using Pointwise Mutual Information
This work addresses the problem of generating summaries without labeled data for users in domains like news and medicine, but it is incremental as it builds on existing unsupervised approaches with a new metric.
The paper tackled unsupervised extractive summarization by proposing new metrics using pointwise mutual information (PMI) to measure sentence relevance and redundancy, and developed a greedy selection algorithm that outperformed similarity-based methods across news, medical, and anecdotal datasets.
Unsupervised approaches to extractive summarization usually rely on a notion of sentence importance defined by the semantic similarity between a sentence and the document. We propose new metrics of relevance and redundancy using pointwise mutual information (PMI) between sentences, which can be easily computed by a pre-trained language model. Intuitively, a relevant sentence allows readers to infer the document content (high PMI with the document), and a redundant sentence can be inferred from the summary (high PMI with the summary). We then develop a greedy sentence selection algorithm to maximize relevance and minimize redundancy of extracted sentences. We show that our method outperforms similarity-based methods on datasets in a range of domains including news, medical journal articles, and personal anecdotes.