I have a project based on tweets wherein I am trying to build a binary classifier, I am aware that I can use a contextual LSTM model which takes the metadata of a tweet as an auxiliary input within the LSTM model. Are there any such models that are versatile to handle textual, numeric and categorical data (one hot encoded/dummy encoded) to classify without creating sparse text matrices.

Ex data:

text                        |topic          | url count      | sentiment
good show, lovely portrayal | topic 1       | 5              | Positive
brutal...wasted time        | topic 2       | 0              | Negative
  • You need to change the textual fields to numerical one by using any encoding. I don't think there is any which can handle textual form. – Rishi Bansal Dec 7 at 10:34
  • @RishiBansal I have vectorized the text by using GLoVE embeddings – aastha Dec 7 at 10:47
  • Tf-idf lasso linear regression can also include other variables than words – keiv.fly Dec 9 at 2:47

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.