LiSum: Open Source Software License Summarization with Multi-Task Learning

TL;DR

This paper introduces LiSum, a multi-task learning approach for automated open source license summarization and classification, supported by a new high-quality dataset and demonstrating significant performance improvements.

Contribution

The paper presents the first automated license summarization method using multi-task learning, along with a new dataset and comprehensive experiments showing superior results.

Findings

LiSum outperforms state-of-the-art baselines in license summarization and classification.

The multi-task learning approach improves F1 scores by at least 5 points.

The authors released datasets, code, and questionnaires for community use.

Abstract

Open source software (OSS) licenses regulate the conditions under which users can reuse, modify, and distribute the software legally. However, there exist various OSS licenses in the community, written in a formal language, which are typically long and complicated to understand. In this paper, we conducted a 661-participants online survey to investigate the perspectives and practices of developers towards OSS licenses. The user study revealed an indeed need for an automated tool to facilitate license understanding. Motivated by the user study and the fast growth of licenses in the community, we propose the first study towards automated license summarization. Specifically, we released the first high quality text summarization dataset and designed two tasks, i.e., license text summarization (LTS), aiming at generating a relatively short summary for an arbitrary license, and license term classification (LTC), focusing on the attitude inference towards a predefined set of key license terms (e.g., Distribute). Aiming at the two tasks, we present LiSum, a multi-task learning method to help developers overcome the obstacles of understanding OSS licenses. Comprehensive experiments demonstrated that the proposed jointly training objective boosted the performance on both tasks, surpassing state-of-the-art baselines with gains of at least 5 points w.r.t. F1 scores of four summarization metrics and achieving 95.13% micro average F1 score for classification simultaneously. We released all the datasets, the replication package, and the questionnaires for the community.