Team Hitachi @ AutoMin 2021: Reference-free Automatic Minuting Pipeline with Argument Structure Construction over Topic-based Summarization

TL;DR

This paper presents a reference-free automatic minuting system that segments transcripts by topics, summarizes them with a fine-tuned BART model, and uses argument mining for coherent structuring, achieving top performance in AutoMin 2021.

Contribution

The novel approach combines topic-based segmentation, pre-trained summarization, and argument mining for structured, coherent minutes without relying on reference data.

Findings

Achieved the best adequacy score in Task A

Outperformed baseline in Tasks B and C

Close to the best system in grammatical correctness

Abstract

This paper introduces the proposed automatic minuting system of the Hitachi team for the First Shared Task on Automatic Minuting (AutoMin-2021). We utilize a reference-free approach (i.e., without using training minutes) for automatic minuting (Task A), which first splits a transcript into blocks on the basis of topics and subsequently summarizes those blocks with a pre-trained BART model fine-tuned on a summarization corpus of chat dialogue. In addition, we apply a technique of argument mining to the generated minutes, reorganizing them in a well-structured and coherent way. We utilize multiple relevance scores to determine whether or not a minute is derived from the same meeting when either a transcript or another minute is given (Task B and C). On top of those scores, we train a conventional machine learning model to bind them and to make final decisions. Consequently, our approach for Task A achieve the best adequacy score among all submissions and close performance to the best system in terms of grammatical correctness and fluency. For Task B and C, the proposed model successfully outperformed a majority vote baseline.