Weak supervision for automatic labeling for my dataset (Deep learning and NLP)

I need weak supervision approach for automatic labeling for my dataset. The problem is, I have labeled dataset with this feature text, verb,prep,start,end, label. For example weve had to break it down into what I think is askable about happy(text),

break(verb),down(prep),(15),(28)(1). The label is 1 for positive or true, and 0 for negative or false. Next I have huge unlabeled text dataset. The weak supervision model (or any else) have to label this dataset for example it has to extract So we take a large collection of images and we break them down into their little image patches(text), break(verb),down(prep),(48),(63)(1) for positive and When organic material dies in nature microbes and bacteria break it and shut down into nutrient rich soil completing the life cycle(text) break(verb),down(prep),(48),(63)(0) for negative. The model should have accuracy, at least 50 %, noisy labeling still be accepting, as long it able to extract and labeling.

Lenguajes naturales Python

Nº del proyecto: #17477244

Sobre el proyecto