Meltem Aksoy
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Meltem Aksoy has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Annika Bush
Smithsonian Tropical Research Institute
2 shared publications Mehmet Fatih Amasyali
Computer Engineering Department, Yıldız Technical University, İstanbul, Turkey
2 shared publications- Mehmet Fatih Amasyali
Yildiz Technical University
2 shared publications Seda Yanık
Industrial Engineering Department, Istanbul Technical University, Istanbul, Turkey
2 shared publications- Seda Yanık
Istanbul Technical University
2 shared publications - Alberto Rigenti
Marketing and Sales Department, Automobili Lamborghini S.p.A., SantAgata Bolognese, Italy
1 shared publication
Latest Publications
- Open Access
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-23
Campus AI vs. Commercial AI: Comparing How Students and Employees Perceive their University’s LLM Chatbot vs. ChatGPT
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-23
Open AccessCampus AI vs. Commercial AI: Comparing How Students and Employees Perceive their University’s LLM Chatbot vs. ChatGPT
- Open Access
- 18 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 43351-43369
Residual Value Prediction in Automotive: A Review of Methods and Data With Research Roadmap
Residual value prediction (RVP) has become a cornerstone of automotive economics, shaping leasing, resale, insurance, fleet management, and long-term brand positioning. Inaccurate predictions can lead to severe financial and strategic consequences: undervaluation undermines competitiveness, while overvaluation erodes profitability and trust. Artificial intelligence (AI) has opened new avenues for enhancing predictive accuracy, giving rise to diverse approaches that leverage heterogeneous data sources, including vehicle condition, ownership history, and macroeconomic factors, alongside advanced modelling techniques and varying levels of transparency and reliability. Given this diversity, a comprehensive synthesis of the field is needed to evaluate current practices and outline future research directions that integrate methodological rigour, multidimensional data inclusion, and principles of trustworthy RVP. This study offers a comprehensive synthesis of the literature by organising existing work into four themes: (i) data-related factors, (ii) prediction and modelling methodologies, (iii) evaluation approaches, and (iv) mechanisms for transparency and trustworthiness. Our proposed research roadmap is guided by responsible AI principles, emphasising the importance of explanations that are both interpretable and reliable for RVP. By consolidating prior contributions and identifying open research opportunities, this study seeks to advance the development of RVP systems that are accurate, transparent, and reliable, thereby supporting more trustworthy decision-making across the automotive ecosystem.
- 18 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 43351-43369
Open AccessResidual Value Prediction in Automotive: A Review of Methods and Data With Research Roadmap
Residual value prediction (RVP) has become a cornerstone of automotive economics, shaping leasing, resale, insurance, fleet management, and long-term brand positioning. Inaccurate predictions can lead to severe financial and strategic consequences: undervaluation undermines competitiveness, while overvaluation erodes profitability and trust. Artificial intelligence (AI) has opened new avenues for enhancing predictive accuracy, giving rise to diverse approaches that leverage heterogeneous data sources, including vehicle condition, ownership history, and macroeconomic factors, alongside advanced modelling techniques and varying levels of transparency and reliability. Given this diversity, a comprehensive synthesis of the field is needed to evaluate current practices and outline future research directions that integrate methodological rigour, multidimensional data inclusion, and principles of trustworthy RVP. This study offers a comprehensive synthesis of the literature by organising existing work into four themes: (i) data-related factors, (ii) prediction and modelling methodologies, (iii) evaluation approaches, and (iv) mechanisms for transparency and trustworthiness. Our proposed research roadmap is guided by responsible AI principles, emphasising the importance of explanations that are both interpretable and reliable for RVP. By consolidating prior contributions and identifying open research opportunities, this study seeks to advance the development of RVP systems that are accurate, transparent, and reliable, thereby supporting more trustworthy decision-making across the automotive ecosystem.
- Open Access
- 3 March 2026
- Published by Wiley in Applied Stochastic Models in Business and Industry
- Vol. 42 (2)
Evaluating Biases in Large Language Models Over Time: A Framework With a GPT Case Study on Political Bias
Large Language Models (LLMs) have repeatedly been shown to reflect systematic biases. At the same time, commercial LLMs are updated at a rapid rate, often without notice to end‐users, so that a bias profile captured today may already be outdated tomorrow. However, the literature still leans heavily on one‐shot evaluations of single model versions, leaving a gap in our understanding of how biases evolve over time and how they should be monitored. We address this gap by introducing a framework for longitudinal evaluation of biases in LLMs, focusing on political bias as a case study. The framework is model‐agnostic, reproducible, and user‐friendly. It consists of (i) locking model versions via dated identifiers to guarantee temporal comparability, (ii) multi‐prompt questionnaires on position statements to analyze potential biases; and (iii) a longitudinal statistical evaluation framework that quantifies and infers absolute bias and drifts between models. Moreover, we suggest conducting (iv) cross‐questionnaire correlation analyses to reveal orthogonal biases, as well as (v) sensitivity analyses on the model's role‐assignment mechanisms to analyze robustness to concrete instructions. All code, prompts, and outputs are openly available to facilitate replication and extension to other bias analyses. To illustrate the framework, we investigate the political biases and personality traits of ChatGPT, specifically comparing GPT‐3.5, GPT‐4, GPT‐4o, and GPT‐5.2. In addition, the ability of the models to emulate political viewpoints (e.g., liberal or conservative positions) is analyzed. Across 4000 generated answers, we observe clear political shifts between versions: While newer models appear less left‐leaning, they still mimic progressive personality profiles and exhibit biases. These findings demonstrate the persistence and transformation of biases across updates, underlining the need for longitudinal monitoring.
- 3 March 2026
- Published by Wiley in Applied Stochastic Models in Business and Industry
- Vol. 42 (2)
Open AccessEvaluating Biases in Large Language Models Over Time: A Framework With a GPT Case Study on Political Bias
Large Language Models (LLMs) have repeatedly been shown to reflect systematic biases. At the same time, commercial LLMs are updated at a rapid rate, often without notice to end‐users, so that a bias profile captured today may already be outdated tomorrow. However, the literature still leans heavily on one‐shot evaluations of single model versions, leaving a gap in our understanding of how biases evolve over time and how they should be monitored. We address this gap by introducing a framework for longitudinal evaluation of biases in LLMs, focusing on political bias as a case study. The framework is model‐agnostic, reproducible, and user‐friendly. It consists of (i) locking model versions via dated identifiers to guarantee temporal comparability, (ii) multi‐prompt questionnaires on position statements to analyze potential biases; and (iii) a longitudinal statistical evaluation framework that quantifies and infers absolute bias and drifts between models. Moreover, we suggest conducting (iv) cross‐questionnaire correlation analyses to reveal orthogonal biases, as well as (v) sensitivity analyses on the model's role‐assignment mechanisms to analyze robustness to concrete instructions. All code, prompts, and outputs are openly available to facilitate replication and extension to other bias analyses. To illustrate the framework, we investigate the political biases and personality traits of ChatGPT, specifically comparing GPT‐3.5, GPT‐4, GPT‐4o, and GPT‐5.2. In addition, the ability of the models to emulate political viewpoints (e.g., liberal or conservative positions) is analyzed. Across 4000 generated answers, we observe clear political shifts between versions: While newer models appear less left‐leaning, they still mimic progressive personality profiles and exhibit biases. These findings demonstrate the persistence and transformation of biases across updates, underlining the need for longitudinal monitoring.
- Open Access
- 12 January 2026
- Published by Wiley in European Journal of Education
- Vol. 61 (1)
Cross‐Cultural Patterns in Artificial Intelligence Literacy Development: Investigating Country and Gender Differences Among Pre‐Service Teachers in Germany and Türkiye
As artificial intelligence (AI) rapidly transforms educational landscapes worldwide, developing AI literacy among pre‐service teachers has become increasingly crucial for effective classroom integration and responsible implementation. This comparative study examines AI literacy levels among pre‐service teachers in Türkiye (Turkey officially changed its name to Türkiye in a 2022 letter to the UN, with the change taking immediate effect. See here ) and Germany, addressing a critical gap in cross‐cultural educational research. Using the Scale for the Assessment of Non‐Experts' AI Literacy (SNAIL), we analysed data from 605 participants through correlation, ANOVA and MANOVA methods. Our findings reveal significant effects of both country and gender on AI literacy development, with German participants demonstrating higher overall literacy scores than their Turkish counterparts, particularly in critical appraisal and practical application dimensions. Conversely, Turkish participants exhibited stronger technical understanding. Across both countries, male participants consistently achieved higher AI literacy scores than females, though this gender gap was most pronounced in technical understanding. The country variable emerged as the more influential factor across all subdimensions. These results highlight the interplay between educational systems, cultural contexts and gender on AI literacy acquisition, offering valuable insights for teacher education programs seeking to prepare educators for increasingly AI‐integrated learning environments.
- 12 January 2026
- Published by Wiley in European Journal of Education
- Vol. 61 (1)
Open AccessCross‐Cultural Patterns in Artificial Intelligence Literacy Development: Investigating Country and Gender Differences Among Pre‐Service Teachers in Germany and Türkiye
As artificial intelligence (AI) rapidly transforms educational landscapes worldwide, developing AI literacy among pre‐service teachers has become increasingly crucial for effective classroom integration and responsible implementation. This comparative study examines AI literacy levels among pre‐service teachers in Türkiye (Turkey officially changed its name to Türkiye in a 2022 letter to the UN, with the change taking immediate effect. See here ) and Germany, addressing a critical gap in cross‐cultural educational research. Using the Scale for the Assessment of Non‐Experts' AI Literacy (SNAIL), we analysed data from 605 participants through correlation, ANOVA and MANOVA methods. Our findings reveal significant effects of both country and gender on AI literacy development, with German participants demonstrating higher overall literacy scores than their Turkish counterparts, particularly in critical appraisal and practical application dimensions. Conversely, Turkish participants exhibited stronger technical understanding. Across both countries, male participants consistently achieved higher AI literacy scores than females, though this gender gap was most pronounced in technical understanding. The country variable emerged as the more influential factor across all subdimensions. These results highlight the interplay between educational systems, cultural contexts and gender on AI literacy acquisition, offering valuable insights for teacher education programs seeking to prepare educators for increasingly AI‐integrated learning environments.
- Open Access
- 23 September 2025
- Published by Elsevier in Intelligent Systems with Applications
- Vol. 28
A data-driven optimization approach for automated reviewer assignment using natural language processing
- 23 September 2025
- Published by Elsevier in Intelligent Systems with Applications
- Vol. 28
Open AccessA data-driven optimization approach for automated reviewer assignment using natural language processing
- Open Access
- 22 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 168183-168201
Trustworthiness Evaluation of Large Language Models Using Multi-Criteria Decision Making
As Large language models (LLMs) become increasingly integrated into high-stakes applications, ensuring their trustworthiness has emerged as a critical research concern. This study proposes a novel evaluation framework that applies a multi-criteria decision making (MCDM) methodology, specifically the hesitant fuzzy analytic hierarchy process (AHP), to assess and rank LLMs based on five key trust dimensions: fairness, robustness, integrity, explainability, and safety. Drawing from expert evaluations, the framework systematically determines the relative importance of each criterion and applies a weighted scoring approach to compare seven leading LLMs, including both proprietary models such as GPT-3.5, GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 and open-source models such as Llama 3.1, Mistral Large 2 and DeepSeek V3. Results reveal GPT-4o as the most trustworthy model, significantly outperforming its peers, particularly in robustness and fairness. Open-source models showed lower scores, especially in safety and explainability, highlighting persistent gaps in their alignment with trust expectations. The findings demonstrate the effectiveness of MCDM in capturing expert uncertainty and prioritizing trust criteria, offering a robust and adaptable framework for evaluating LLMs in dynamic and sensitive domains.
- 22 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 168183-168201
Open AccessTrustworthiness Evaluation of Large Language Models Using Multi-Criteria Decision Making
As Large language models (LLMs) become increasingly integrated into high-stakes applications, ensuring their trustworthiness has emerged as a critical research concern. This study proposes a novel evaluation framework that applies a multi-criteria decision making (MCDM) methodology, specifically the hesitant fuzzy analytic hierarchy process (AHP), to assess and rank LLMs based on five key trust dimensions: fairness, robustness, integrity, explainability, and safety. Drawing from expert evaluations, the framework systematically determines the relative importance of each criterion and applies a weighted scoring approach to compare seven leading LLMs, including both proprietary models such as GPT-3.5, GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 and open-source models such as Llama 3.1, Mistral Large 2 and DeepSeek V3. Results reveal GPT-4o as the most trustworthy model, significantly outperforming its peers, particularly in robustness and fairness. Open-source models showed lower scores, especially in safety and explainability, highlighting persistent gaps in their alignment with trust expectations. The findings demonstrate the effectiveness of MCDM in capturing expert uncertainty and prioritizing trust criteria, offering a robust and adaptable framework for evaluating LLMs in dynamic and sensitive domains.
- Top Cited PapersOpen Access
- 2 July 2025
- Published by Elsevier in Natural Language Processing Journal
- Vol. 12
Whose morality do they speak? Unraveling cultural bias in multilingual language models
- 2 July 2025
- Published by Elsevier in Natural Language Processing Journal
- Vol. 12
Top Cited PapersOpen AccessWhose morality do they speak? Unraveling cultural bias in multilingual language models
- Open Access
Reviewer Assignment Problem: A Systematic Review of the Literature
Appropriate reviewer assignment significantly impacts the quality of proposal evaluation, as accurate and fair reviews are contingent on their assignment to relevant reviewers. The crucial task of assigning reviewers to submitted proposals is the starting point of the review process and is also known as the reviewer assignment problem (RAP). Due to the obvious restrictions of manual assignment, journal editors, conference organizers, and grant managers demand automatic reviewer assignment approaches. Many studies have proposed assignment solutions in response to the demand for automated procedures since 1992. The primary objective of this survey paper is to provide scholars and practitioners with a comprehensive overview of available research on the RAP. To achieve this goal, this article presents an in-depth systematic review of 103 publications in the field of reviewer assignment published in the past three decades and available in the Web of Science, Scopus, ScienceDirect, Google Scholar, and Semantic Scholar databases. This review paper classified and discussed the RAP approaches into two broad categories and numerous subcategories based on their underlying techniques. Furthermore, potential future research directions for each category are presented. This survey shows that the research on the RAP is becoming more significant and that more effort is required to develop new approaches and a framework.
Open AccessReviewer Assignment Problem: A Systematic Review of the Literature
Appropriate reviewer assignment significantly impacts the quality of proposal evaluation, as accurate and fair reviews are contingent on their assignment to relevant reviewers. The crucial task of assigning reviewers to submitted proposals is the starting point of the review process and is also known as the reviewer assignment problem (RAP). Due to the obvious restrictions of manual assignment, journal editors, conference organizers, and grant managers demand automatic reviewer assignment approaches. Many studies have proposed assignment solutions in response to the demand for automated procedures since 1992. The primary objective of this survey paper is to provide scholars and practitioners with a comprehensive overview of available research on the RAP. To achieve this goal, this article presents an in-depth systematic review of 103 publications in the field of reviewer assignment published in the past three decades and available in the Web of Science, Scopus, ScienceDirect, Google Scholar, and Semantic Scholar databases. This review paper classified and discussed the RAP approaches into two broad categories and numerous subcategories based on their underlying techniques. Furthermore, potential future research directions for each category are presented. This survey shows that the research on the RAP is becoming more significant and that more effort is required to develop new approaches and a framework.
- 28 February 2023
- Published by Emerald Publishing in International Journal of Intelligent Computing and Cybernetics
- Vol. 16 (3) , 595-628
A comparative analysis of text representation, classification and clustering methods over real project proposals
Purpose: When a large number of project proposals are evaluated to allocate available funds, grouping them based on their similarities is beneficial. Current approaches to group proposals are primarily based on manual matching of similar topics, discipline areas and keywords declared by project applicants. When the number of proposals increases, this task becomes complex and requires excessive time. This paper aims to demonstrate how to effectively use the rich information in the titles and abstracts of Turkish project proposals to group them automatically.Design/methodology/approach: This study proposes a model that effectively groups Turkish project proposals by combining word embedding, clustering and classification techniques. The proposed model uses FastText, BERT and term frequency/inverse document frequency (TF/IDF) word-embedding techniques to extract terms from the titles and abstracts of project proposals in Turkish. The extracted terms were grouped using both the clustering and classification techniques. Natural groups contained within the corpus were discovered using k-means, k-means++, k-medoids and agglomerative clustering algorithms. Additionally, this study employs classification approaches to predict the target class for each document in the corpus. To classify project proposals, various classifiers, including k-nearest neighbors (KNN), support vector machines (SVM), artificial neural networks (ANN), classification and regression trees (CART) and random forest (RF), are used. Empirical experiments were conducted to validate the effectiveness of the proposed method by using real data from the Istanbul Development Agency.Findings: The results show that the generated word embeddings can effectively represent proposal texts as vectors, and can be used as inputs for clustering or classification algorithms. Using clustering algorithms, the document corpus is divided into five groups. In addition, the results demonstrate that the proposals can easily be categorized into predefined categories using classification algorithms. SVM-Linear achieved the highest prediction accuracy (89.2%) with the FastText word embedding method. A comparison of manual grouping with automatic classification and clustering results revealed that both classification and clustering techniques have a high success rate.Research limitations/implications: The proposed model automatically benefits from the rich information in project proposals and significantly reduces numerous time-consuming tasks that managers must perform manually. Thus, it eliminates the drawbacks of the current manual methods and yields significantly more accurate results. In the future, additional experiments should be conducted to validate the proposed method using data from other funding organizations.Originality/value: This study presents the application of word embedding methods to effectively use the rich information in the titles and abstracts of Turkish project proposals. Existing research studies focus on the automatic grouping of proposals; traditional frequency-based word embedding methods are used for feature extraction methods to represent project proposals. Unlike previous research, this study employs two outperforming neural network-based textual feature extraction techniques to obtain terms representing the proposals: BERT as a contextual word embedding method and FastText as a static word embedding method. Moreover, to the best of our knowledge, there has been no research conducted on the grouping of project proposals in Turkish.
- 28 February 2023
- Published by Emerald Publishing in International Journal of Intelligent Computing and Cybernetics
- Vol. 16 (3) , 595-628
A comparative analysis of text representation, classification and clustering methods over real project proposals
Purpose: When a large number of project proposals are evaluated to allocate available funds, grouping them based on their similarities is beneficial. Current approaches to group proposals are primarily based on manual matching of similar topics, discipline areas and keywords declared by project applicants. When the number of proposals increases, this task becomes complex and requires excessive time. This paper aims to demonstrate how to effectively use the rich information in the titles and abstracts of Turkish project proposals to group them automatically.Design/methodology/approach: This study proposes a model that effectively groups Turkish project proposals by combining word embedding, clustering and classification techniques. The proposed model uses FastText, BERT and term frequency/inverse document frequency (TF/IDF) word-embedding techniques to extract terms from the titles and abstracts of project proposals in Turkish. The extracted terms were grouped using both the clustering and classification techniques. Natural groups contained within the corpus were discovered using k-means, k-means++, k-medoids and agglomerative clustering algorithms. Additionally, this study employs classification approaches to predict the target class for each document in the corpus. To classify project proposals, various classifiers, including k-nearest neighbors (KNN), support vector machines (SVM), artificial neural networks (ANN), classification and regression trees (CART) and random forest (RF), are used. Empirical experiments were conducted to validate the effectiveness of the proposed method by using real data from the Istanbul Development Agency.Findings: The results show that the generated word embeddings can effectively represent proposal texts as vectors, and can be used as inputs for clustering or classification algorithms. Using clustering algorithms, the document corpus is divided into five groups. In addition, the results demonstrate that the proposals can easily be categorized into predefined categories using classification algorithms. SVM-Linear achieved the highest prediction accuracy (89.2%) with the FastText word embedding method. A comparison of manual grouping with automatic classification and clustering results revealed that both classification and clustering techniques have a high success rate.Research limitations/implications: The proposed model automatically benefits from the rich information in project proposals and significantly reduces numerous time-consuming tasks that managers must perform manually. Thus, it eliminates the drawbacks of the current manual methods and yields significantly more accurate results. In the future, additional experiments should be conducted to validate the proposed method using data from other funding organizations.Originality/value: This study presents the application of word embedding methods to effectively use the rich information in the titles and abstracts of Turkish project proposals. Existing research studies focus on the automatic grouping of proposals; traditional frequency-based word embedding methods are used for feature extraction methods to represent project proposals. Unlike previous research, this study employs two outperforming neural network-based textual feature extraction techniques to obtain terms representing the proposals: BERT as a contextual word embedding method and FastText as a static word embedding method. Moreover, to the best of our knowledge, there has been no research conducted on the grouping of project proposals in Turkish.
- 5 July 2022
- Published by Springer Nature
- p. 990-998
Intelligent Word Embedding Methods to Support Project Proposal Grouping for Project Selection
- 5 July 2022
- Published by Springer Nature
- p. 990-998
Intelligent Word Embedding Methods to Support Project Proposal Grouping for Project Selection