Araştırma Makalesi

KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES

Sayı: 2 11 Haziran 2026
PDF İndir
EN TR

KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES

Öz

Background/Aim: The growing number of students in higher education institutions has intensified the demand for efficient and accessible information services. Traditional methods such as call centers, physical help desks, and static FAQ pages are inherently limited in scalability, availability, and personalization. This study aims to present KÜBİT AI, a domainspecific chatbot system designed for Kırıkkale University that combines a RetrievalAugmented Generation (RAG) architecture with a locally deployed large language model (LLM). Methods: The proposed system integrates a multilingual sentence-transformer-based semantic retrieval engine over a structured knowledge base compiled from official Kırıkkale University domains with a DuckDuckGo-based web-search fallback mechanism, supporting both structured knowledge access and real-time information discovery. The backend is built upon a Flask RESTful API and communicates with an Ollama-hosted OpenAI GPT-OSS (20B) model. Evaluation is conducted using the RAGAS framework. The study further reports a comparison against BM25, TF-IDF, and pure-LLM baselines, an ablation over retrieval thresholds and top-k, a stress test on a low-overlap paraphrased question set, an evaluation of the web-fallback component, and an error-taxonomy-based failure analysis. Results: The main system obtains context precision of 0.918, context recall of 0.987, faithfulness of 0.782, answer relevancy of 0.630, and answer correctness of 0.778, reaching a RAGAS harmonic mean of 0.805. Conclusion: The core RAG pipeline is designed to run locally, with only the optional websearch fallback relying on an external service when up-to-date public information is needed. This design contributes to data privacy and institutional autonomy

Anahtar Kelimeler

Kaynakça

  1. Følstad, A., Brandtzaeg, P. B. 2017. Chatbots and the New World of HCI. Interactions, 24(4), 38-42.
  2. Adamopoulou, E., Moussiades, L. 2020. Chatbots: History, Technology, and Applications. Machine Learning with Applications, 2, 100006.
  3. Zhao, W. X., Zhou, K., Li, J., Tang, T., Wang, X., Hou, Y., et al. 2023. A Survey of Large Language Models. https://arxiv.org/abs/2303.18223
  4. Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., et al. 2023. Survey of Hallucination in Natural Language Generation. ACM Computing Surveys, 55(12), 1-38.
  5. Huang, L., Yu, W., Ma, W., Zhong, W., Feng, Z., Wang, H., et al. 2024. A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions. https://arxiv.org/abs/2311.05232
  6. Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., et al. 2020. RetrievalAugmented Generation for Knowledge-Intensive NLP Tasks. Advances in Neural Information Processing Systems, 33, 9459-9474.
  7. Gao, Y., Xiong, Y., Gao, X., Jia, K., Pan, J., Bi, Y., et al. 2024. Retrieval-Augmented Generation for Large Language Models: A Survey. https://arxiv.org/abs/2312.10997
  8. Khasanova Zafar kizi, M., Suh, Y. 2025. Design and Performance Evaluation of LLMBased RAG Pipelines for Chatbot Services in International Student Admissions. Electronics, 14(15), 3095.

Ayrıntılar

Birincil Dil

İngilizce

Konular

Yazılım Mühendisliği (Diğer)

Bölüm

Araştırma Makalesi

Yayımlanma Tarihi

11 Haziran 2026

Gönderilme Tarihi

27 Şubat 2026

Kabul Tarihi

4 Mayıs 2026

Yayımlandığı Sayı

Yıl 2026 Sayı: 2

Kaynak Göster

APA
Balcı, M., Erten, M. Y., & Aydilek, H. (2026). KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES. Enginoscope, 2. https://izlik.org/JA47EL25AH
AMA
1.Balcı M, Erten MY, Aydilek H. KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES. Enginoscope. 2026;(2). https://izlik.org/JA47EL25AH
Chicago
Balcı, Musa, Mustafa Yasin Erten, ve Hüseyin Aydilek. 2026. “KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES”. Enginoscope, sy 2. https://izlik.org/JA47EL25AH.
EndNote
Balcı M, Erten MY, Aydilek H (01 Haziran 2026) KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES. Enginoscope 2
IEEE
[1]M. Balcı, M. Y. Erten, ve H. Aydilek, “KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES”, Enginoscope, sy 2, Haz. 2026, [çevrimiçi]. Erişim adresi: https://izlik.org/JA47EL25AH
ISNAD
Balcı, Musa - Erten, Mustafa Yasin - Aydilek, Hüseyin. “KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES”. Enginoscope. 2 (01 Haziran 2026). https://izlik.org/JA47EL25AH.
JAMA
1.Balcı M, Erten MY, Aydilek H. KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES. Enginoscope. 2026. Available at https://izlik.org/JA47EL25AH.
MLA
Balcı, Musa, vd. “KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES”. Enginoscope, sy 2, Haziran 2026, https://izlik.org/JA47EL25AH.
Vancouver
1.Musa Balcı, Mustafa Yasin Erten, Hüseyin Aydilek. KUBIT AI: A RETRIEVAL AUGMENTED GENERATION BASED INTELLIGENT CHATBOT SYSTEM FOR KIRIKKALE UNIVERSTY STUDENT SERVICES. Enginoscope [Internet]. 01 Haziran 2026;(2). Erişim adresi: https://izlik.org/JA47EL25AH