Automatic Generation of Indicative Multidocument Summaries

Title: Automatic Generation of Indicative Multidocument Summaries

Research Question: How can natural language generation be used to create indicative multidocument summaries that help users decide which documents to read?

Methodology: The researchers analyzed the indicative summarization task from a generation perspective. They first studied the required content by examining published guidelines and corpus analysis. They then showed how these summaries can be factored into a set of document features and how an implemented content planner uses the topicality document feature to create indicative multidocument query-based summaries.

Results: The study found that indicative summaries can be improved by focusing on document features such as topic distribution and media type. They developed a natural language generation model that automatically creates indicative multidocument summaries based on these features.

Implications: The research suggests that natural language generation can be used to create more effective indicative multidocument summaries, which can help users make better decisions about which documents to read. This could have significant implications for information retrieval and user experience in search engines and other information systems.

Link to Article: https://arxiv.org/abs/0107019v1 Authors: arXiv ID: 0107019v1