The Practical Challenges of Active Learning: Lessons Learned from Live Experimentation
This work addresses practical implementation issues of active learning for researchers and practitioners in natural language processing, particularly for low-resource languages like Thai, but it is incremental as it focuses on lessons learned rather than new methods.
The study tested active learning for selecting Thai text sentences for human annotation in a live setting, comparing it with random sampling, and described practical challenges and interactions with environmental changes encountered during the experiment.
We tested in a live setting the use of active learning for selecting text sentences for human annotations used in training a Thai segmentation machine learning model. In our study, two concurrent annotated samples were constructed, one through random sampling of sentences from a text corpus, and the other through model-based scoring and ranking of sentences from the same corpus. In the course of the experiment, we observed the effect of significant changes to the learning environment which are likely to occur in real-world learning tasks. We describe how our active learning strategy interacted with these events and discuss other practical challenges encountered in using active learning in the live setting.