Expert-Workshop: Transforming Data into Insight: Data Annotation in Machine Learning for Environmental Research and Forestry

The rapid development of artificial intelligence (AI) methods is opening up ground-breaking opportunities to analyse and find solutions to the most pressing socio-political challenges of our time. AI methods have the potential to significantly advance and change research in the fields of environmental protection and sustainability (BMUV, 2023). The results must not only be reliable but also comprehensible. This lends credibility to the solutions proposed. In addition to explainable or interpretable models, the availability of high-quality annotated data is essential for the models to learn how to correctly recognise the phenomenon behind the observations.

Annotated data form the basis for good AI-based modelling and serve as a driving force for the further development of AI-driven environmental research. Data accuracy and quality are the decisive criteria for the result, even more than its quantity. It is often said that 80% of the work of a ML practitioner goes in data processing. It is often pointed out that most of the effort involved in building an ML-based model must be invested in annotation and other types of data pre-processing. Recognising data gaps, identifying areas of application with high demand, and developing efficient annotation strategies are building blocks for the successful and effective use of AI in the environmental sector.

Within this expert-Workshop, will give an overview of the problem of data annotation in environmental research and development, and then focus on concrete problems faced by researchers in forestry. Are there similar issues or do they differ from the rest in characteristics such as their format; accessibility; quality? We will discover it during the workshop on July 30 from 10:00 to 11:30.

Transforming Data into Insight: Data Annotation in Machine Learning for Environmental Research and Forestry

July 30 from 10:00 to 11:30