I have a dataset similar to newsgroup20 for classification. With the training dataset, I have a dictionary data set that explains some jargons in the training dataset. These both are different data set, So how will i utilize the dictionary dataset for improving my model accuracy?
You can try this approach LEARNING TO COMPUTE WORD EMBEDDINGS ON THE FLY. Using additional auxiliary data you can substitute jargons using definitions from dictionary. In this case, jargons embeddings will become more relevant, in semantic sense .