Text biases
By now, you should recognize the patterns for fetching real-world image datasets and importing metadata into pandas. It is the same pattern for text datasets. Pluto will guide you through two sessions and use his power of observation to name the biases. He could employ the latest in generative AI such as OpenAI GPT3 or GPT4 to list the biases in the text. Maybe he will do that later, but for now, he will use his noggin. Nevertheless, Pluto will attempt to write Python code to gain insight into the texts' structures, such as the word count and misspelled words. It is not the fairness matrix but a step in the right direction.
Pluto searches the Kaggle website for the Natural Language Processing (NLP) dataset, and the result consists of over 2,000 datasets. He chooses the Netflix Shows and the Amazon Reviews datasets. Retrieving and viewing the NLP dataset follows the same fetching, importing, and printing steps outlined in the image dataset.
Let’s start...
