[论文解读] Leveraging BERT for Extractive Text Summarization on Lectures

Derek M. Miller|arXiv (Cornell University)|Jun 7, 2019

Topic Modeling参考文献 14被引用 196

一句话总结

论文提出 Lecture Summarization Service，一个 Python RESTful 系统，使用 BERT 嵌入和 KMeans 聚类从讲座文字记录中提取关键句，供用户指定的摘要长度。

ABSTRACT

In the last two decades, automatic extractive text summarization on lectures has demonstrated to be a useful tool for collecting key phrases and sentences that best represent the content. However, many current approaches utilize dated approaches, producing sub-par outputs or requiring several hours of manual tuning to produce meaningful results. Recently, new machine learning architectures have provided mechanisms for extractive summarization through the clustering of output embeddings from deep learning models. This paper reports on the project called Lecture Summarization Service, a python based RESTful service that utilizes the BERT model for text embeddings and KMeans clustering to identify sentences closes to the centroid for summary selection. The purpose of the service was to provide students a utility that could summarize lecture content, based on their desired number of sentences. On top of the summary work, the service also includes lecture and summary management, storing content on the cloud which can be used for collaboration. While the results of utilizing BERT for extractive summarization were promising, there were still areas where the model struggled, providing feature research opportunities for further improvement.

研究动机与目标

• 激发并实现对讲座内容的自动化抽取式摘要。
• 利用 BERT 生成有意义的句子嵌入以用于基于聚类的摘要。
• 提供一个可扩展的云端讲座摘要与内容管理服务。

提出的方法

• 使用 BERT 模型计算文本嵌入以表示句子。
• 对句子嵌入应用 KMeans 聚类以形成簇。
• 选择最接近簇质心的句子来形成摘要。
• 提供一个基于 Python 的 RESTful 服务以生成用户自定义长度的摘要。
• 包括用于讲座和摘要管理的功能，并带有用于协作的云存储。

实验结果

研究问题

• RQ1 基于 BERT 的句子嵌入能否提升讲座抽取式摘要的质量？
• RQ2 基于聚类的选择（靠近质心的句子）是否能有效生成简明、具有代表性的讲座内容摘要？
• RQ3 云端支持的 RESTful 服务是否能合理支持用户自定义摘要长度和协作管理？

主要发现

• BERT 嵌入 + KMeans 聚类可以识别用于摘要的代表性句子。
• 该服务提供了一个可实际操作的工作流程，用于生成用户自定义长度的摘要。
• 该方法显示出有前景的结果，但也存在模型遇到困难的领域和改进的机会。
• 该系统包括讲座与摘要管理，提供云存储以支持协作。