Examining the State-of-the-Art in News Timeline Summarization
This work addresses the problem of evaluating and advancing timeline summarization for news analysis, though it appears incremental as it builds on existing methods.
The paper tackled the unclear state-of-the-art in news timeline summarization by comparing strategies and proposing a simple combined method, which improved over SOTA on all tested benchmarks, and introduced a larger, longer-spanning dataset for more robust evaluation.
Previous work on automatic news timeline summarization (TLS) leaves an unclear picture about how this task can generally be approached and how well it is currently solved. This is mostly due to the focus on individual subtasks, such as date selection and date summarization, and to the previous lack of appropriate evaluation metrics for the full TLS task. In this paper, we compare different TLS strategies using appropriate evaluation frameworks, and propose a simple and effective combination of methods that improves over the state-of-the-art on all tested benchmarks. For a more robust evaluation, we also present a new TLS dataset, which is larger and spans longer time periods than previous datasets.