I want to build a deep learning algorithm that places emojis in a pretty long text, please don't ask me why I just want to do it. Where should I start? I've read that recurrent neural networks are good for stuff involving text. What to do next?

  • $\begingroup$ It's going to be difficult to answer this question without a few more details. Is the placement of emojis difficult to do programmatically? Should the emojis be related to the nearby text? Do you have training data (i.e. samples of text where emojis are appropriately placed)? $\endgroup$ – zachdj Dec 2 '19 at 17:47
  • 1
    $\begingroup$ @zachdj first of all thanks for replying; I am sorry for the lack of details; since the emojis are related to the nearby text I think it is difficult to guess programmatically since it implies an understanding of the text. Also I have not specified but it needs to guess which emoji to use. I have training data: i have samples of text where they are correctly placed. $\endgroup$ – Solal Baudoin Dec 2 '19 at 20:53
  • $\begingroup$ If you have enough samples, then deep learning could work for this problem. Look into sequence-to-sequence models for text. These models take a sequence as input and produce a sequence as output. In your case, the input sequence would be the text without emojis, and the output sequence would be the text with emojis sprinkled in. $\endgroup$ – zachdj Dec 3 '19 at 17:06
  • $\begingroup$ I would recommend treating each emoji as a "word", then training word embeddings on your text corpus. Each word will be represented by a vector, and your model will accept a sequence of such vectors. $\endgroup$ – zachdj Dec 3 '19 at 17:09
  • $\begingroup$ @zachdj thank you so much i'll update you on that^^ $\endgroup$ – Solal Baudoin Dec 3 '19 at 17:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.