The Difference between Extraction and Classification Models
The aim of this article is to compare Extraction and Classification models within KAPTO. While these models share similarities, their primary objectives and training processes differ significantly. The article explores how extraction models identify entities within a document, while classification models aim to categorise the document into specific classes.
While Extraction and Classification Models share many similarities, there are also significant differences.
1. One of the most critical differences is their primary objective. Extraction models aim to detect entities within a document, while Classification models aim to divide the document into categories.
2. Another difference is the training process. As we have seen, extraction models require the labelling of specific entities within the data, while classification models require the labelling of the data into categories. This difference in the training process leads to differences in the output of the models.
3. A classification model's output typically categorises the data into specific classes, while the output of an extraction model is a structured representation of the entities within the data.
