ELEFANT

Children’s language skills are an important foundation for a successful school career and participation in society. However, many children already show difficulties learning to read when they start school—for example, due to language development disorders or because they are acquiring German as a second language in preschool. In preschools, educational professionals therefore face the challenge of identifying individual support needs early on and providing appropriate assistance, despite limited staff resources. The goal: to enable every child, as far as possible, to make a smooth transition from preschool to school.

The “ELEFANT” project addresses this challenge by developing an AI-supported digital support program. Using automated language analysis and adaptive learning activities, the project aims to specifically prepare children in their final year of preschool for learning to read without placing a significant additional workload on the daily routine of the preschool.

The goal of the project is to scientifically investigate the effectiveness of digital support. At the same time, the project aims to develop a system that assists early childhood educators in observing and documenting children’s language development and support processes.

The project consortium is developing a corresponding training approach and testing it in real-world settings with preschool groups using tablets. Over several months, the children listen to and watch digitally presented stories during regular sessions, complete individual exercises, and then retell the stories. Their spoken language is analyzed using automatic speech recognition and AI-based methods. Based on this analysis, the difficulty level of the exercises can be individually adjusted. The focus is on promoting phonological awareness—that is, the conscious handling of sounds and syllables—as well as narrative skills, which encompass understanding and retelling stories.

The results of the children participating in the digital reading support program are compared with those of children who receive either traditional read-aloud support or the regular support provided as part of their daily routine at daycare.

Tasks of the project partners

The Fraunhofer IDMT in Oldenburg coordinates the “ELEFANT” project and contributes its expertise in the automatic analysis of children’s speech as well as the development of digital applications suitable for everyday use. This includes, in particular, the technical and AI-supported implementation of speech recognition and analysis, as well as the adaptive adjustment of learning activities.

LMU Munich complements the consortium with scientific expertise in the areas of early reading support and educational psychology. It helps to provide a scientific foundation for the support content, assist with the conduct of the study, and scientifically evaluate the program’s effectiveness.

Both project partners have already gained experience in the field of digital reading applications. Together, they aim to help make language support during the transition from preschool to school more individualized, effective, and suitable for everyday use.

Legal notice: The software demonstrators mentioned here are intended exclusively for research and development purposes (research use only). The software is designed as part of a system and is not intended to be used as a standalone medical device.