I am having difficulty using BERT for a sentiment analysis task that handles both aspect-based sentiment analysis (ABSA) and comment sentiment analysis. I know that using two separate classification layers on top of the hidden layer can classify these two tasks, but I don't know how to handle labels for these two datasets. One dataset only has sentiment labels (positive, negative, and neutral), while the other includes 10 aspects with 4 sentiment labels (including NONE for aspects that are not present). I am wondering if I can assign a task type to the data and use it to identify which task is being trained on the model. Please let me know your thoughts on this problem. Thank you for your help! Have a good day!!!!!



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.