Text Classification

[MathBart]

Hi !

I have a dataset of school comments. I am looking for a text classifier which could be able to classify them in 2 categories : efficient or not. For example :

• “The work is pretty good overall. Keep going” → not efficient
• “Timothée’s foundations are fragile and pose difficulties for him in individual written productions. However, in class, he is capable of much better (especially orally) where his interventions and his exchanges with his classmates show much better skills. I encourage him in his efforts. He is capable of progress.” → efficient

Any advice ?

Thank you in advance

[nielsr]

Hi,

Sure that’s possible, refer to the docs: Text classification.

To load your local dataset, refer to this guide: Load text data.

[KateWinslet]

@MathBart

I have a dataset of school comments. I am looking for a text classifier which could be able to classify them in 2 categories : efficient or not. For example :

• “The work is pretty good overall. Keep going” → not efficient
• “Timothée’s foundations are fragile and pose difficulties for him in individual written productions. However, in class, he is capable of much better (especially orally) where his interventions and his exchanges with his classmates show much better skills. I encourage him in his efforts. He is capable of progress.” → efficient

Any advice ?

You should Utilize NLP techniques with supervised learning. Preprocess, vectorize, and train classifiers like SVM, Naive Bayes on labeled data for efficient/non-efficient categorization.