arXiv subjects

Mahmudur Rashid

Publications and source records attributed to Mahmudur Rashid.

3 recordsLinked to original sources

arXiv2026

Seeing Without Understanding: Large Language Model Evaluation of Mobile User Interface Quality, Failure Taxonomy, and Architectural Explanation

Evaluating mobile user interface quality at scale remains a persistent challenge in software engineering and human-computer interaction. Rule-based heuristic methods offer structural reliability but demand significant engineering effort, while human annotation does not scale to the volume of applications produced annually. Large language models present a promising alternative, yet their reliability for structured UI judgment has not been systematically examined, and the patterns behind their failures remain insufficiently characterized. This paper addresses both gaps. We begin with the complete RICO dataset of 66,261 real-world mobile application screens, from which we derive a refined evaluation corpus of 15,000 screens through a rigorous, literature-guided selection process. Each screen is assessed across seven criteria: structural JSON validity, minimum visible element count, clickable component presence, non-zero layout bounds, image integrity, and perceptual duplicate removal. Against this corpus, we apply a heuristic baseline built from severity-weighted usability signals, normalized layout metrics, and pixel-ratio complexity measures calibrated to real user sentiment. Multiple language models independently rate each screen across usability, layout quality, and visual complexity from structured JSON descriptions and raw screenshots. Dimension-level comparison against the heuristic uses agreement rates, Cohen's Kappa, and confidence calibration. Recurring divergence patterns are organized into a failure taxonomy and interpreted through transformer architectural signatures: MLE plausibility bias, attention misgrounding, and autoregressive over-commitment.

cs.HC↗

arXiv2021

Reinforcing Iron Metal Matrix Composite by Multi-Wall Carbon Nanotube: A Combined Theoretical and Computational Approach

Carbon nanotube (CNT) reinforced metal matrix composites have been the focus of researchers due to their high load-bearing capacity. Among single and multi-wall carbon nanotubes (MWCNT), the latter is preferred by manufacturers and engineers for making composites due to their economic feasibility of synthesizing. However, the effect of layer numbers along with other parameters of the reinforcing MWCNT must be understood before its industrial application. In this article, we developed a novel theoretical approach for predicting the variation of strength and stiffness of MWCNT reinforced iron composites (MWCNT-Fe) with the layer number of reinforcing MWCNT and validated the prediction with a series of Molecular dynamics (MD) simulation. Our analysis revealed that for every addition of two extra layers, the strength and stiffness of the composite increase 9.8% and 7.2% respectively up to eight layered MWCNT and then becomes saturated. We also employed MD simulations for investigating the effect of grain boundary on the failure mechanism of CNT reinforced iron composites in contrast to previous studies. Our investigations revealed that instead of the matrix-fiber interface, the failure was initiated from the grain boundary and merges with the interface. The results in this study will not only help engineers and manufacturers choose optimal layered MWCNT for synthesizing composite for a specific application but also provide scientists a new method to model composites for predicting desired properties.

cond-mat.mtrl-sci↗

arXiv2020

A Molecular Dynamics Investigation of Mechanical Properties of Graphene Reinforced Iron Composite and The Effect of Vacancy Defect Distance from the Matrix-Fiber Interface

Graphene is a material of excellent mechanical properties, which make it an ideal fiber for reinforcing metal. Since iron is the most used metal in the world, reinforcing iron with graphene can reduce the overall requirement of material in any application where strength is demanded. However, the effect of graphene reinforcement on the mechanical properties of iron needs to be known before the industrial application of the composite. In this paper, we have investigated the mechanical properties of graphene-reinforced iron composite by Molecular Dynamics (MD) method for various conditions. The properties were investigated by applying uniaxial tension on a modeled representative volume element (RVE). The effect of temperature on the mechanical property of the composite was also studied because the knowledge is required for manufacturing products with the composite operating at a wide temperature range. MD analysis also revealed that the initiation of fracture is from the matrix-fiber interface. We also investigated how the distance of vacancy defects from the matrix-fiber interface affects the mechanical properties of the composite, which can be used to select a suitable manufacturing process. The results obtained from this study show that vacancy defects lower the strength at a greater extent as it gets closer to the interface.

cond-mat.mtrl-sci↗