The goal of the project is to develop quality metrics for Slovene primary and secondary school textbooks to be used during their certification and evaluation. The research group will develop a prototype of the automatic web-based tool, which will allow evaluation of textbooks based on their textual and pictorial properties. The tool will support decision-making process for selection of appropriate textbook.

UL FRI researchers will adapt, develop and implement several readability metrics for the Slovene language. They will use techniques from natural language processing, machine learning, and image recognition to assess readability and text coherence. The decision support will be implemented through a hierarchical multi-parameter decision model, which will be flexible enough to capture requirements of different areas (social, humanistic and natural sciences), different age groups, education levels, special needs of pupils etc. The developed tool will be calibrated and evaluated on a sample of human evaluated textbooks. The tool will be open-sourced and publicly available and will provide feedback on readability and text coherence.