CORDIAL develops innovative approaches for integrating Artificial Intelligence (AI) into corpus-based teaching and research in English linguistics. The project centers on the development of CorpusMentor, an AI-supported assistant that will be integrated into the BonnCorpora platform and designed to support students throughout the corpus-linguistic research process.

 Corpus linguistics is a central methodology in contemporary language studies, yet many students encounter substantial technical barriers when working with large language datasets, corpus query languages, annotation systems, and statistical analyses. CORDIAL addresses these challenges by combining corpus technology with natural-language interfaces and adaptive AI support.

 At the core of the project is a locally hosted, open-source large language model (LLM) that enables users to interact with corpora through natural language. Importantly, the degree of AI assistance is configurable by instructors and can be aligned with specific learning objectives. In some courses, CorpusMentor may simply help students formulate valid corpus queries by translating natural-language requests into Corpus Query Language (CQP) syntax. In other contexts, instructors may choose to enable more advanced functionality, such as guided corpus exploration, automated visualizations, methodological explanations, or support in interpreting corpus evidence.

 This flexible design allows CorpusMentor to function either as a low-level technical assistant or as a more comprehensive learning companion. Students can gradually develop corpus-linguistic skills while maintaining transparency regarding the underlying analytical procedures. Rather than replacing methodological training, the system is intended to scaffold learning and support the development of independent research competencies.

 The project also contributes to the continued development of the BonnCorpora infrastructure. During the funding period, the platform will be expanded with standardized corpus formats, advanced corpus search technologies, interactive visualization tools, and AI-assisted learning components. Particular emphasis is placed on creating reusable, sustainable, and privacy-conscious solutions that can be deployed locally and adapted to different teaching scenarios.

 By combining corpus linguistics, artificial intelligence, and inquiry-based learning, CORDIAL aims to lower barriers to empirical language research while preserving methodological rigor. The project seeks to establish a flexible framework that enables instructors to determine how and to what extent AI is integrated into corpus-based teaching and learning.

Project members:

Prof. Dr. Robert Fuchs (PI)

Julia Schilling, M.Ed. (Technical Development & Research Integration)