Shujaat Ali

Shujaat Ali

Auto-generated profile, based on publicly available information

Shujaat Ali

Bacha Khan University Charsadda Pakistan

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 April 2026 in Engineering Reports
Verified
Journal Article
Open access
MJMalik Asif Jilani
Malik Asif Jilani
INIftikhar Azim Niaz
Iftikhar Azim Niaz
SZSaad Naeem Zafar
Saad Naeem Zafar
FSFaiz Ali Shah
Faiz Ali Shah
SAShujaat Ali
Shujaat Ali
MTMuhammad Tahir
Muhammad Tahir
Usability is a fundamental software quality attribute that strongly influences user performance, acceptance, and the overall success of software systems. Although Dromey's Quality Model provides an important relationship between high‐level quality attributes and low‐level structural forms through quality‐carrying properties (QCPs), it does not explicitly provide a usability‐focused mechanism that can guide developers and quality engineers in systematically incorporating usability into software products. Existing quality models often treat usability at a broad or abstract level, making it difficult to trace usability concerns to concrete software structures and diagnose usability defects systematically. This creates a gap between high‐level usability goals and practical software evaluation and development activities. To address this gap, this study proposes the usability quality model (UQM) as an enhancement of Dromey's Model. The model was developed by identifying relevant structural forms, extracting usability carrying properties (UCPs) from established usability guidelines and literature, organizing usability into sub‐attributes, and mapping these sub‐attributes to associated high‐level usability attributes. Usability defects were also incorporated to show how the absence or violation of required UCPs can reduce software usability. The proposed model provides a structured linkage among usability attributes, sub‐attributes, UCPs, structural forms, and usability defects. As a proof of concept, the model was applied to common usability defects identified in the literature, demonstrating how usability problems can be traced to structural forms and how appropriate UCPs can guide corrective actions. The results show that UQM can support both top‐down and bottom‐up usability analysis. The proposed UQM extends Dromey's Model into a more usability‐oriented and practically interpretable framework for software quality assessment. It can assist developers, quality managers, and evaluators in embedding usability into software products and in systematically identifying and addressing usability defects. The model also provides a foundation for future empirical validation, real‐world case studies, and automation support for usability‐aware software quality evaluation.

1 January 2026 in Complexity
Verified
Journal Article
Open access
PDF file
RSRahat Ali Shah
Rahat Ali Shah
JUJamal Uddin
Jamal Uddin
HHHameed Hussain
Hameed Hussain
KKKhaistah Khan
Khaistah Khan
SAShujaat Ali
Shujaat Ali
MTMuhammad Tahir
Muhammad Tahir
Frequent itemset mining (FIM) is a fundamental task in data mining with applications ranging from market basket analysis and recommendation systems to fraud detection and bioinformatics. However, mining frequent itemsets from massive datasets remains computationally challenging due to high execution time and memory consumption. Traditional algorithms such as Apriori suffer from exponential candidate generation and repeated database scans, FP‐Growth incurs high memory overhead from tree construction, and Equivalence Class Transformation (ECLAT) faces expensive tid‐list intersections. To address these limitations, we propose RPriori, a novel recursive divide‐and‐conquer algorithm that operates on the item‐space instead of the transaction‐space. RPriori recursively partitions the set of items and terminates early when a subset is found frequent, thereby eliminating candidate generation and avoiding global data structures. It performs localized support counting and merges only verified frequent subsets, resulting in O (NMlogM) time complexity and O (M) space complexity. Extensive experiments on six benchmark datasets, namely, Bakery, Chess, Mushroom, Retail, Accident, and T10I4D100K, demonstrate that RPriori consistently outperforms Apriori, FP‐Growth, ECLAT, and TL‐DIC. At low support thresholds, RPriori achieves up to 91% reduction in runtime and 93% reduction in memory usage, confirming its superior scalability and efficiency for both sparse and dense datasets.

1 January 2026 in Journal of Electrical and Computer Engineering
Verified
Journal Article
Open access
TKTabassum Kanwal
Tabassum Kanwal
RYRehan Mehmood Yousaf
Rehan Mehmood Yousaf
Saud AltafSaud Altaf
Saud Altaf
SAShujaat Ali
Shujaat Ali
DSDilawar Shah
Dilawar Shah
MTMuhammad Tahir
Muhammad Tahir
Sign language is a primary communication modality for many deaf and hard‐of‐hearing individuals, yet many recognition systems remain limited by static‐gesture assumptions and weak generalization to realistic acquisition variability. This paper presents a real‐time dynamic Pakistan Sign Language (PSL) gesture recognition framework that combines YOLOv8 for hand localization with recurrent temporal modeling for sequence classification. A dynamic PSL dataset was collected from 20 participants covering 20 gesture classes (2220 RGB and Kinect video sequences), recorded across five sessions, multiple environments, three lighting conditions (bright, moderate, and dim), and three capture distances (2, 3, and 4 ft). To address generalization concerns, we report signer‐independent leave‐one‐subject‐out (LOSO) evaluation with mean ± standard deviation, along with complementary cross‐session and cross‐environment tests, and an aggregate confusion matrix computed over the independent test predictions. Under LOSO, the proposed temporal model using bounding box and velocity features achieves 93.04% ± 1.61% accuracy with macro‐F1 of 0.930 ± 0.016. Cross‐session testing (train sessions 1–4 and test session 5) yields 95.62% ± 0.84% accuracy, while cross‐environment testing (leave‐one‐environment‐out) achieves 91.18% ± 1.94% accuracy, quantifying performance under domain shift. We further provide controlled baselines and ablations, including YOLO‐only majority voting, CNN + LSTM on YOLO‐cropped hand images, a lightweight 3D CNN on short clips, and a GRU variant, confirming the contribution of temporal modeling and motion cues. The system maintains accuracy above 90% under varying lighting and distance conditions, supporting practical real‐time deployment. Limitations include the current focus on manual hand gestures only; future work will integrate nonmanual cues (face and body pose) for more linguistically complete PSL processing.

1 January 2026 in Applied Computational Intelligence and Soft Computing
Verified
Journal Article
Open access
PDF file
MAMuhammad Atif
Muhammad Atif
IKIzaz Ahmad Khan
Izaz Ahmad Khan
Mohammad AbrarMohammad Abrar
Mohammad Abrar
SAShujaat Ali
Shujaat Ali
DSDilawar Shah
Dilawar Shah
MTMuhammad Tahir
Muhammad Tahir
Magnetic resonance imaging (MRI) is widely used for brain‐tumor assessment, but conventional deep learning workflows often require patient images to be centralized, creating privacy, governance, and regulatory barriers. This study presents a lightweight federated learning (FL) framework for brain MRI classification using a multilayer perceptron (MLP) trained with the Flower framework and Federated Averaging (FedAvg). In the revised formulation, the task is treated strictly as classification: each MRI‐derived sample is represented by a 239‐dimensional feature vector and assigned to one of the encoded diagnostic classes in the label column. Segmentation terminology has been removed or clarified, and the Dice coefficient is reported only as a class‐level F1/Dice‐style classification metric rather than as a pixel‐level tumor‐overlap score. The prepared BraTS‐derived subset used for simulation contains 3000 .npy samples organized for training, normal/abnormal representation, and mask‐associated metadata; the full BraTS source and the preprocessed subset are now distinguished to avoid confusion. The FL setting is explicitly described as a simulation in which a single prepared dataset is partitioned into artificial clients, not as a true multihospital deployment. The proposed MLP contains 39,814 trainable parameters, requiring approximately 0.152 MB per model update in a 32‐bit floating‐point format, which makes it substantially cheaper to communicate than typical CNN‐based alternatives. The best reported FL–MLP configuration achieved 95% accuracy, 0.92 class‐level Dice/F1, 0.91 sensitivity, and 0.97 specificity. The revised manuscript also clarifies that FL reduces the need for raw data sharing but does not provide formal privacy guarantees unless combined with secure aggregation, differential privacy, encryption, or attack‐resistance mechanisms.

4 June 2026 in Journal of Software: Evolution and Process
Verified
Journal Article
Closed access
PDF file
FKFarhan Mateen Khan
Farhan Mateen Khan
AMAsim Munir
Asim Munir
SAShujaat Ali
Shujaat Ali
TJTahir Jameel
Tahir Jameel
MKMohammed Asmat Ullah Khan
Mohammed Asmat Ullah Khan
MTMuhammad Tahir
Muhammad Tahir