salience

In short, a graph-based ranking algorithm is a way of deciding on the importance of a vertex within a graph, by taking into account global information recursively computed from the entire graph, rather than relying only on local vertex-specific information ... In this paper, we introduced TextRank – a graph-based ranking model for text processing.

(Rada Mihalcea and Paul Tarau, "TextRank: Bringing Order into Texts", 2004)

Extractive summarization should be preferred over abstractive summarization when nuance is essential and when the summary is meant as a companion to the source text. LLMs are effective at abstractive summarization, but they can also be leveraged in extractive summarization. Rather than solve this with a prompt, LLM embeddings can be combined with the TextRank algorithm to reliably yield high-quality extractive summaries.

In other words, embeddings can be used to automatically generate highlights. The internal representation is an affinity matrix between sentences that can be used to find the most salient sentences in a text.

Preview

How to Run

$ poetry install
$ poetry run flask --app salience run

This will kick off a flask server. You can access the output at http://127.0.0.1:5000/static/index.html