Information Systems: Computing Sciences: Knowledge Systems: Intel-Miner: Semantic Machine for Robust Interpretation of Noisy Intelligence
Abstract
Multi-Perspective Narratives (MPN) are ubiquitous, useful tools for verifying facts from different alternative narratives; thus, MPNs facilitate more informed decisions by offering a concise overall picture of the current situation. Assimilation and digestion of such MPNs at a large scale pose significant challenges for users,making it extremely arduous to filter out reliable information from these MPNs. Despite great progress made in natural language processing (NLP), computers are still far from being able to accurately analyze multiperspective narratives and effectively summarize the common information/overlapping content they provide. In this project, we will design and develop solutions for this crucial yet relatively unexplored NLP task, which we formally refer to as Semantic Overlap Summarization (SOS). SOS entails generating a single summary from multiple alternative narratives that can convey the common information provided.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 31, 2023
- Accession Number
- AD1226939
Entities
People
- Shubhra K. Karmaker
Organizations
- Auburn University