NOVA introduces a multi-stage system combining belief elicitation techniques and narrative structure designs to facilitate users in assessing their perception of mainstream media coverage. The system allows users to compare their personal beliefs with data-driven visualizations, encouraging critical thinking and evidence-based assessment.
The content discusses the challenges of assessing media bias and the importance of externalizing personal beliefs. It highlights the need for user-friendly tools like NOVA to guide individuals in evaluating news media coverage objectively.
Recent developments in Natural Language Processing have sparked an increased effort to rethink how we analyze media bias. Social science and psychology researchers have been calling for an interdisciplinary approach with computer science to assess news media content. NOVA aims to address these issues by providing a visual interface for users to evaluate their personal beliefs on media coverage biases.
NOVA's design considerations focus on supporting sensemaking, externalizing personal beliefs, comparing personal beliefs with data, and evaluating personal beliefs through evidence finding. The system is structured into three stages: Topic Selection, Belief Elicitation, and Article Review, allowing users to navigate through complex media coverage data effectively.
The study design involved recruiting participants from the United States to evaluate NOVA's effectiveness in assisting individuals in assessing their personal beliefs on news media coverage. Participants engaged in tasks that required them to compare their perceptions with data-driven visualizations and find evidence to support or disprove their beliefs.
Sang ngôn ngữ khác
từ nội dung nguồn
arxiv.org
Thông tin chi tiết chính được chắt lọc từ
by Keshav Dasu,... lúc arxiv.org 03-04-2024
https://arxiv.org/pdf/2403.00334.pdfYêu cầu sâu hơn