Rebwar Mala Nabi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Rebwar Mala Nabi has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Rania Azad M. San Ahmed
Sulaimani Polytechnic University
3 shared publications - Payman Sabr Rostam
Department of Information Technology, Computer Science Institute, Sulaimani Polytechnic University, Sulaimani, Kurdistan Region, Iraq
2 shared publications - Sardasht M-Raouf Mahmood
Statistics and Computer Dept, University of Sulaimani
2 shared publications - Chovyan H. Wahid
Computer Department, College of Science, Charmo University, Sulaymaniyah, Iraq
1 shared publication - Dana Lattef Hussein
IT Department, Computer Science Institute, Sulaimani Polytechnic University, Sulaymaniyah, Kurdistan Region, Iraq
1 shared publication Habibolah Bin Harron
1 shared publication
Latest Publications
- 24 September 2026
- Published by Springer Nature in International Journal of Data Science and Analytics
- Vol. 22 (1) , 310
Towards robust stance detection: data augmentation, model fine-tuning, and empirical benchmarking
- 24 September 2026
- Published by Springer Nature in International Journal of Data Science and Analytics
- Vol. 22 (1) , 310
Towards robust stance detection: data augmentation, model fine-tuning, and empirical benchmarking
- Open Access
- 17 July 2026
- Published by Koya University in ARO-THE SCIENTIFIC JOURNAL OF KOYA UNIVERSITY
- Vol. 14 (2) , 18-27
The First Comprehensive Study of Stance Detection Modeling for the Sorani Kurdish Language
Stance detection has become a fundamental task in natural language processing (NLP), yet it remains under-explored for low-resource languages such as Sorani Kurdish. Building on the previously released Bochun dataset, the present work focuses exclusively on the comprehensive evaluation of stance detection models and provides the first benchmarking of classical machine learning, deep learning (DL), and transformers. Eight models are implemented: Support vector machine (SVM), logistic regression, random forest, extreme gradient boosting, convolutional neural network, bidirectional long short-term memory, and two frozen transformer encoders, Central Kurdish Bidirectional Encoder Representations from Transformers (BERT) and XLM-RoBERTa-base (XLM-R), each combined with a logistic-regression classification head. All models are trained and evaluated under four protocols (80/20 and 70/30 stratified splits and 5-fold and 10-fold cross-validation) with five random seeds. Class-imbalance handling is investigated through a dedicated ablation comparing no weighting, model-appropriate weighting, and Synthetic Minority Oversampling Technique oversampling, and pairwise statistical comparisons are conducted using Wilcoxon signed-rank tests complemented by Cohen’s d effect sizes. With the chosen feature representations and dataset size, the SVM trained on term frequency–inverse document frequency features delivers the strongest performance (accuracy and weighted F1 of 74% on the 80/20 split and 72% under 10-fold cross-validation), outperforming the DL and frozen-transformer baselines. The findings provide a foundational reference point for future research on Kurdish NLP in general and stance detection specifically.
- 17 July 2026
- Published by Koya University in ARO-THE SCIENTIFIC JOURNAL OF KOYA UNIVERSITY
- Vol. 14 (2) , 18-27
Open AccessThe First Comprehensive Study of Stance Detection Modeling for the Sorani Kurdish Language
Stance detection has become a fundamental task in natural language processing (NLP), yet it remains under-explored for low-resource languages such as Sorani Kurdish. Building on the previously released Bochun dataset, the present work focuses exclusively on the comprehensive evaluation of stance detection models and provides the first benchmarking of classical machine learning, deep learning (DL), and transformers. Eight models are implemented: Support vector machine (SVM), logistic regression, random forest, extreme gradient boosting, convolutional neural network, bidirectional long short-term memory, and two frozen transformer encoders, Central Kurdish Bidirectional Encoder Representations from Transformers (BERT) and XLM-RoBERTa-base (XLM-R), each combined with a logistic-regression classification head. All models are trained and evaluated under four protocols (80/20 and 70/30 stratified splits and 5-fold and 10-fold cross-validation) with five random seeds. Class-imbalance handling is investigated through a dedicated ablation comparing no weighting, model-appropriate weighting, and Synthetic Minority Oversampling Technique oversampling, and pairwise statistical comparisons are conducted using Wilcoxon signed-rank tests complemented by Cohen’s d effect sizes. With the chosen feature representations and dataset size, the SVM trained on term frequency–inverse document frequency features delivers the strongest performance (accuracy and weighted F1 of 74% on the 80/20 split and 72% under 10-fold cross-validation), outperforming the DL and frozen-transformer baselines. The findings provide a foundational reference point for future research on Kurdish NLP in general and stance detection specifically.
- Open Access
- 30 October 2025
- Published by Elsevier in Data in Brief
- Vol. 63, 112215
KuSarcasm: Automated annotation of a sarcasm dataset using hybrid NLP techniques
- 30 October 2025
- Published by Elsevier in Data in Brief
- Vol. 63, 112215
Open AccessKuSarcasm: Automated annotation of a sarcasm dataset using hybrid NLP techniques
- Open Access
- 22 August 2025
- Published by Elsevier in Data in Brief
- Vol. 62, 111999
Adyan: automated annotating named entity recognition dataset for sorani kurdish language
- 22 August 2025
- Published by Elsevier in Data in Brief
- Vol. 62, 111999
Open AccessAdyan: automated annotating named entity recognition dataset for sorani kurdish language
- Open Access
- 25 June 2025
- Published by Elsevier in Data in Brief
- Vol. 61, 111839
Bochun: Automatically annotated stance detection dataset for Sorani Kurdish language
- 25 June 2025
- Published by Elsevier in Data in Brief
- Vol. 61, 111839
Open AccessBochun: Automatically annotated stance detection dataset for Sorani Kurdish language
- Open Access
- 1 January 2023
- Published by Sulaimani Polytechnic University in Kurdistan Journal of Applied Research
- Vol. 8 (1) , 73-85
Enhanced AdaBoostM1 with Multilayer Perceptron for Stock Price Prediction
Stock market investment has gained significant popularity due to its potential for economic returns, prompting extensive research in financial time series forecasting. Among the predictive models, various adaptations of the AdaBoostM1 algorithm have been applied to stock market prediction, either by tuning parameters or experimenting with different base learners. However, the achieved accuracy often remains suboptimal. This study addresses these limitations by introducing an enhanced version of AdaBoostM1 (ADA), implemented on the Waikato Environment for Knowledge Analysis (WEKA) platform, to forecast stock prices using historical data. The proposed model, termed AdaBoost with Multilayer Perceptron (ADA-MLP), replaces the commonly used Decision stumps with a set of Multilayer Perceptron (MLP) models as weak learners. The experimental results demonstrate that ADA-MLP consistently outperformed the standard AdaBoostM1 algorithm, achieving an average classification accuracy of 100%, compared to 98.48% by AdaBoostM1—a relative improvement of 1.52%. Additionally, ADA-MLP demonstrated superior performance against other enhanced versions of AdaBoost presented in prior studies, achieving an average of 5.3% higher accuracy. Statistical significance testing using the paired t-test confirmed the reliability of these results, with p-values < 0.05. The experiments were conducted on the Yahoo finance dataset from 25 years of historical data spanning from January 1995 to January 2020, comprising 6295 samples, ensuring a robust and comprehensive evaluation. These findings highlight the potential of ADA-MLP to enhance financial forecasting and offer a reliable tool for stock market prediction. Future research could explore extending this approach to other financial instruments and larger datasets to further validate its effectiveness.
- 1 January 2023
- Published by Sulaimani Polytechnic University in Kurdistan Journal of Applied Research
- Vol. 8 (1) , 73-85
Open AccessEnhanced AdaBoostM1 with Multilayer Perceptron for Stock Price Prediction
Stock market investment has gained significant popularity due to its potential for economic returns, prompting extensive research in financial time series forecasting. Among the predictive models, various adaptations of the AdaBoostM1 algorithm have been applied to stock market prediction, either by tuning parameters or experimenting with different base learners. However, the achieved accuracy often remains suboptimal. This study addresses these limitations by introducing an enhanced version of AdaBoostM1 (ADA), implemented on the Waikato Environment for Knowledge Analysis (WEKA) platform, to forecast stock prices using historical data. The proposed model, termed AdaBoost with Multilayer Perceptron (ADA-MLP), replaces the commonly used Decision stumps with a set of Multilayer Perceptron (MLP) models as weak learners. The experimental results demonstrate that ADA-MLP consistently outperformed the standard AdaBoostM1 algorithm, achieving an average classification accuracy of 100%, compared to 98.48% by AdaBoostM1—a relative improvement of 1.52%. Additionally, ADA-MLP demonstrated superior performance against other enhanced versions of AdaBoost presented in prior studies, achieving an average of 5.3% higher accuracy. Statistical significance testing using the paired t-test confirmed the reliability of these results, with p-values < 0.05. The experiments were conducted on the Yahoo finance dataset from 25 years of historical data spanning from January 1995 to January 2020, comprising 6295 samples, ensuring a robust and comprehensive evaluation. These findings highlight the potential of ADA-MLP to enhance financial forecasting and offer a reliable tool for stock market prediction. Future research could explore extending this approach to other financial instruments and larger datasets to further validate its effectiveness.
- Open Access
- 1 December 2019
- Published by Science Publications in Journal of Computer Science
- Vol. 15 (12) , 1795-1808
Ultimate Prediction of Stock Market Price Movement
- 1 December 2019
- Published by Science Publications in Journal of Computer Science
- Vol. 15 (12) , 1795-1808
Open AccessUltimate Prediction of Stock Market Price Movement
- Open Access
- 1 January 2018
- Published by Sulaimani Polytechnic University in Kurdistan Journal of Applied Research
- Vol. 3 (1) , 27-33
The Impact of Teaching Materials on Learning Computer Programming Languages in Kurdistan Region Universities and Institutes
It is evident that learning and teaching computer programming are considered as one of the striking challenges in academic environments. Meanwhile, selecting the correct and appropriate materials can leave an enormous impact in learning computer programming languages. However, recently this argument has been put under scrutiny as to which types of materials motivate learners to learn computer programming languages as well as enhance learning outcomes. Therefore, the main objective of this study is to investigate the current teaching and learning materials of computer programming languages in Kurdistan region of Iraq universities. Additionally, another aim is to give a rigorous analysis of how materials help students to learn computer programming language. A further focus is to identify the difficulties of learning computer programming languages at undergraduate level which constitutes technical Diploma and Bachelor. The last but not the least, this paper examines new approaches to teaching programming languages as a cognitive model for programming education.
- 1 January 2018
- Published by Sulaimani Polytechnic University in Kurdistan Journal of Applied Research
- Vol. 3 (1) , 27-33
Open AccessThe Impact of Teaching Materials on Learning Computer Programming Languages in Kurdistan Region Universities and Institutes
It is evident that learning and teaching computer programming are considered as one of the striking challenges in academic environments. Meanwhile, selecting the correct and appropriate materials can leave an enormous impact in learning computer programming languages. However, recently this argument has been put under scrutiny as to which types of materials motivate learners to learn computer programming languages as well as enhance learning outcomes. Therefore, the main objective of this study is to investigate the current teaching and learning materials of computer programming languages in Kurdistan region of Iraq universities. Additionally, another aim is to give a rigorous analysis of how materials help students to learn computer programming language. A further focus is to identify the difficulties of learning computer programming languages at undergraduate level which constitutes technical Diploma and Bachelor. The last but not the least, this paper examines new approaches to teaching programming languages as a cognitive model for programming education.
- Open Access
- 27 August 2017
- Published by Sulaimani Polytechnic University in Kurdistan Journal of Applied Research
- Vol. 2 (3) , 25-31
Optimization Algorithm’s Problems: Comparison Study
Currently, in various fields and disciplines problem optimization are used commonly. In this concern, we have to define solutions which are two known concepts optimal or near optimal optimization problems in regards to some objects. Usually, it is surely difficult to sort problems out in only one step, but some processes can be followed by us which people usually call it problem solving. Frequently, the solution process is split into various steps which are accomplishing one after the other. Therefore, in this paper we consider some algorithms that help us to sort out problems, for exemplify, finding the shortest path, minimum spanning tree, maximum network flows and maximum matching. More importantly, the algorithm comparison will be presented. Additionally, the limitation of each algorithm. The last but not the least, the future research in this area will be approached.
- 27 August 2017
- Published by Sulaimani Polytechnic University in Kurdistan Journal of Applied Research
- Vol. 2 (3) , 25-31
Open AccessOptimization Algorithm’s Problems: Comparison Study
Currently, in various fields and disciplines problem optimization are used commonly. In this concern, we have to define solutions which are two known concepts optimal or near optimal optimization problems in regards to some objects. Usually, it is surely difficult to sort problems out in only one step, but some processes can be followed by us which people usually call it problem solving. Frequently, the solution process is split into various steps which are accomplishing one after the other. Therefore, in this paper we consider some algorithms that help us to sort out problems, for exemplify, finding the shortest path, minimum spanning tree, maximum network flows and maximum matching. More importantly, the algorithm comparison will be presented. Additionally, the limitation of each algorithm. The last but not the least, the future research in this area will be approached.
- 1 December 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 519-525
State-of-the-Art in Handwritten Signature Verification System
Recently, handwritten signature verification (HSV) has become atremendously active area of research. Considerable results have been achieved in terms of accuracy and computation so far. Generally, biometrics can be divided into two types:Behavioral (signature verification, keystroke dynamics, etc.) and Physiological (iris characteristics, fingerprint, etc.). Signature verification is widely studied and discussed by using two approaches, on-line and offline approaches. Offline systems are more applicable and easy to use in comparison with on-line systems in many parts of the world. However, it is considered more difficult than on-line verification due to the lack of dynamic information. This paper presents State-of-Art about both types of HSV Systems. In this paper, we present recent methods used to capture data as well as different methods and techniques used in pre-processing steps. Additionally, current methods used for features extraction and approaches used for verification in signature systems are presented. Finally, we discuss approaches as well as techniques that have been used. In conclusion, we recommend encouraging ideas to be incorporated in the future.
- 1 December 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 519-525
State-of-the-Art in Handwritten Signature Verification System
Recently, handwritten signature verification (HSV) has become atremendously active area of research. Considerable results have been achieved in terms of accuracy and computation so far. Generally, biometrics can be divided into two types:Behavioral (signature verification, keystroke dynamics, etc.) and Physiological (iris characteristics, fingerprint, etc.). Signature verification is widely studied and discussed by using two approaches, on-line and offline approaches. Offline systems are more applicable and easy to use in comparison with on-line systems in many parts of the world. However, it is considered more difficult than on-line verification due to the lack of dynamic information. This paper presents State-of-Art about both types of HSV Systems. In this paper, we present recent methods used to capture data as well as different methods and techniques used in pre-processing steps. Additionally, current methods used for features extraction and approaches used for verification in signature systems are presented. Finally, we discuss approaches as well as techniques that have been used. In conclusion, we recommend encouraging ideas to be incorporated in the future.