Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection
Permanent URI for this collectionhttps://hdl.handle.net/11147/7148
Browse
5 results
Search Results
Article Toward Reliable Annotation in Low-Resource NLP: A Mixture of Agents Framework and Multi-LLM Benchmarking(IEEE-Inst Electrical Electronics Engineers Inc, 2025) Onan, Aytug; Nasution, Arbi Haza; Celikten, TugbaThis paper introduces the Mixture-of-Agents (MoA) framework, a structured approach for improving the reliability of large language model (LLM)-based text annotation in low-resource NLP contexts. MoA employs coordinated agent interactions to enhance agreement, interpretability, and robustness without manual supervision. Evaluations on Turkish classification benchmarks demonstrate that MoA achieves up to 10-point improvements in macro-F1 over single-model baselines and significantly increases inter-agent consistency. Additionally, three novel reliability metrics-Conflict Rate (CR), Ambiguity Resolution Success Rate (ARSR), and Refinement Correction Rate (RCR)-are proposed to quantify annotation stability and correction dynamics. The results indicate that multi-agent coordination can substantially improve label quality, offering a scalable pathway toward trustworthy annotation in low-resource and cross-domain applications. The framework is language-agnostic and adaptable to other low-resource contexts beyond Turkish, including morphologically rich or typologically diverse languages such as Indonesian, Urdu, and Swahili. These findings highlight the scalability of MoA as a generalizable solution for multilingual and cross-domain annotation.Article Automating Software Size Measurement from Python Code Using Language Models(Springer, 2025) Tenekeci, Samet; Unlu, Huseyin; Gul, Bedir Arda; Keles, Damla; Kuuk, Murat; Demirors, OnurSoftware size is a key input for project planning, effort estimation, and productivity analysis. While pre-trained language models have shown promise in deriving functional size from natural-language requirements, measuring size directly from source code remains under-explored. Yet, code-based size measurement is critical in modern workflows where requirement documents are often incomplete or unavailable, especially in Agile development environments. This exploratory study investigates the use of CodeBERT, a pre-trained bimodal transformer model, for measuring software size directly from Python source code according to two measurement methods: COSMIC Function Points and MicroM. We construct two curated datasets from the Python subset of the CodeSearchNet corpus, and manually annotate each function with its corresponding size. Our experimental results show that CodeBERT can successfully measure COSMIC data movements with up to 91.4% accuracy and generalize to the functional, architectural, and algorithmic event types defined in MicroM, reaching up to 81.5% accuracy. These findings highlight the potential of code-based language models for automated functional size measurement when requirement artifacts are absent or unreliable.Article Citation - WoS: 1Citation - Scopus: 1Automating Software Size Measurement With Language Models: Insights From Industrial Case Studies(Elsevier Science Inc, 2026) Unlu, Huseyin; Tenekeci, Samet; Kennouche, Dhia Eddine; Demirors, OnurObjective software size measurement is critical for accurate effort estimation, yet many organizations avoid it due to high costs, required expertise, and time-consuming manual effort. This often leads to vague predictions, poor planning, and project overruns. To address this challenge, we investigate the use of pre-trained language models - BERT and SE-BERT - to automate size measurement based on textual requirements using COSMIC and MicroM methods. We constructed one heterogeneous dataset and two industrial datasets, each manually measured by experienced analysts. Models were evaluated in three settings: (i) generic model evaluation, where the models are trained and tested on heterogeneous data, (ii) internal evaluation, where the models are trained and tested on organization-specific data, and (iii) external evaluation, where generic models were tested on organization-specific data. Results show that organization-specific models significantly outperform generic models, indicating that aligning training data with the target organization's requirement style is critical for accuracy. SE-BERT, a domain-adapted variant of BERT, improves performance, particularly in low-resource settings. These findings highlight the practical potential of tailoring training data for broader adoption and cost-effective software size measurement in industrial contexts.Conference Object Citation - WoS: 3Citation - Scopus: 5Predicting Software Functional Size Using Natural Language Processing: an Exploratory Case Study(IEEE, 2024) Unlu, Huseyin; Tenekeci, Samet; Ciftci, Can; Oral, Ibrahim Baran; Atalay, Tunahan; Hacaloglu, Tuna; Demirors, OnurSoftware Size Measurement (SSM) plays an essential role in software project management as it enables the acquisition of software size, which is the primary input for development effort and schedule estimation. However, many small and medium-sized companies cannot perform objective SSM and Software Effort Estimation (SEE) due to the lack of resources and an expert workforce. This results in inadequate estimates and projects exceeding the planned time and budget. Therefore, organizations need to perform objective SSM and SEE using minimal resources without an expert workforce. In this research, we conducted an exploratory case study to predict the functional size of software project requirements using state-of-the-art large language models (LLMs). For this aim, we fine-tuned BERT and BERT_SE with a set of user stories and their respective functional size in COSMIC Function Points (CFP). We gathered the user stories included in different project requirement documents. In total size prediction, we achieved 72.8% accuracy with BERT and 74.4% accuracy with BERT_SE. In data movement-based size prediction, we achieved 87.5% average accuracy with BERT and 88.1% average accuracy with BERT_SE. Although we use relatively small datasets in model training, these results are promising and hold significant value as they demonstrate the practical utility of language models in SSM.Article Citation - Scopus: 2Turkmednli: a Turkish Medical Natural Language Inference Dataset Through Large Language Model Based Translation(Peerj inc, 2025) Ogul, Iskender Ulgen; Soygazi, Fatih; Bostanoglu, Belgin ErgencNatural language inference (NLI) is a subfield of natural language processing (NLP) that aims to identify the contextual relationship between premise and hypothesis sentences. While high-resource languages like English benefit from robust and rich NLI datasets, creating similar datasets for low-resource languages is challenging due to the cost and complexity of manual annotation. Although translation of existing datasets offers a practical solution, direct translation of domain-specific datasets presents unique challenges, particularly in handling abbreviations, metric conversions, and cultural alignment. This study introduces a pipeline for translating a medical NLI dataset into Turkish, which is a low-resource language. Our approach employs fine-tuning the Llama-3.1 model with selected samples from the Medical Abbreviation dataset (MeDAL) to extract and resolve medical abbreviations. Consequently, NLI pairs are refined with extracted abbreviations and subjected to metric correction. Later, the processed sentences are then translated using Facebook's No Language Left Behind (NLLB) translation model. To ensure quality, we conducted comprehensive evaluations using both machine learning models and medical expert review. Our results show that BERTurk achieved 75.17% accuracy on TurkMedNLI test data and 76.30% on the normalized test set, while BioBERTurk demonstrated comparable performance with 75.59% accuracy on test data and 72.29% on the normalized dataset. Medical experts further validated the translations through manual assessment of sampled sentences. This work demonstrates the effectiveness of large language models in adapting domain-specific datasets for low-resource languages, establishing a foundation for future research in multilingual biomedical NLP.