Annotator bias 💫

There are many reasons why Machine Learning models sometimes do not generalize well in production.

One of them may be hidden in your annotation process 🧵 Annotator bias is a type of data bias that comes from a lack of diversity in the annotator population 👥

The bias can result in low-quality datasets if the data does not fit the intended use case

Models trained on biased datasets do not perform well in production

Annotation guidelines for ML 101 (basic concepts) 🧵

✨ The results of supervised learning approaches are only as good as the annotations they are based on.

✨ Annotations are only as good as the guidelines that annotators rely upon to direct their efforts.

⬇️ What is data annotation? 🪄

✨It is the task of associating entries in your data with additional information

✨It is also known as "coding" or "labeling"

✨It is crucial for both qualitative and quantitative analyses

⬇️