GRAFFAIN is a multi-module, online analysis software powered by Python Shiny that enables qualitative data analysis and visualization with the support of large language models.

While GRAFFAIN aims to increase the reproducibility of qualitative data analysis, it also acknowledges that the nature of qualitative research involves the researcher’s interpretation and theoretical perspective. Therefore, GRAFFAIN's core philosophy is to assist—not replace—researchers.

GRAFFAIN uses multilingual pre-trained language models for data analysis. The strength and performance of each module depend on the capabilities of the language model used. While GRAFFAIN strives to provide consistent results, it accepts that some errors in classification or analysis may occur and constantly aims to minimize them. Please be aware that the likelihood of such errors may vary depending on the language and content of your dataset. For best results, we recommend comparing your findings across different language formats.

Each module in GRAFFAIN is designed for a specific analytical purpose and uses language models trained with relevant prompts and datasets. The module’s performance can increase or decrease depending on the similarity between your dataset and the training data. The amount of error may also vary based on the intended use versus the original training objective of the model.

GRAFFAIN modules are trained for specific, limited purposes and may not deliver reliable results if used outside these intended scopes. Therefore, we recommend that users carefully review and interpret their results.

A detailed technical background report for each module will be published soon in the documentation section of our website. You can access the quick user guide by clicking the button below:

GRAFFAIN is a free and independent platform developed through the dedicated efforts of its creators to support scientific and social research. If you use GRAFFAIN in your work, citing the developers is the only expectation in return for their efforts.

Citation:

Çüm, S., Demir, E. K., Demir, T., & Kahyaoğlu Erdoğmuş, Y. (2025). GRAFFAIN: Mini-Guide[Preprint]. figshare. https://doi.org/10.6084/m9.figshare.29376641

Beyond the specific contributions listed below, the GRAFFAIN founding and management team collaborated closely on the ideas, design, creation, and testing processes. This spirit of collective development will continue as GRAFFAIN evolves through user feedback.

GRAFFAIN welcomes contributions from all researchers. You can contact the developer of any module directly using the contact addresses below to provide feedback or suggestions. The modular structure of GRAFFAIN reflects its vision of continuous growth. Existing modules will be continuously improved and new modules will be added. The management team invites all researchers interested in collaborating and developing new modules to join the platform. Contributors will become part of the GRAFFAIN volunteer team and will be acknowledged as developers of their modules.

Assoc. Prof. Dr. Sait Çüm

Dokuz Eylül University
Educational Measurement and Evaluation
Contact: sait.cum@deu.edu.tr
Modules Developed: Content Analysis, Network Analysis, Sentiment Shift Analysis, Mainstream & Marginal Opinion Analysis, Shiny Design, Shiny UI & Server

Assoc. Prof. Dr. Elif Kübra Demir

Ege University
Educational Measurement and Evaluation
Contact: elif.kubra.demir@ege.edu.tr
Modules Developed: Four Emotion Analysis, Risk Matrix, Shiny UI & Server

Dr. Tolga Demir

Ministry of National Education
Contact: demir.tolga@yahoo.com
Modules Developed: Content Analysis, Needs Assessment, Political Axis Analysis, Shiny UI & Server

Dr. Yasemin Kahyaoğlu Erdoğmuş

Dokuz Eylül University
Computer Education & Instructional Technology
Contact: yasemin.kahyaoglu@deu.edu.tr
Modules Developed: Logo Design, Website Design, Web Hosting, Website Frontend and Backend