2
$\begingroup$

I don't have a background in neural networks. But, various studies has been proved that neural networks (feed forward / Recurrent) outperformed n-gram language modeling for predicting words in a sequence. But, in an application to text messaging or any text-based conversation, where the language which is most likely used will be more informal or colloquial. Can still a neural networks perform well than n-gram LM? Considering the data to be fed are the text messages (colloquial phrases). If so, please enlighten me, thanks.

$\endgroup$

1 Answer 1

4
$\begingroup$

A neural network is in principle a good choice when you have A LOT of similar data and classification tasks. Predicting the next character (or word... which is just multiple characters) is such a szenario. I don't think it really matters which kind of language you have, as long as you have enough training data of the same kind.

See The Unreasonable Effectiveness of Recurrent Neural Networks for a nice article where a recurrent neural network (RNN) was used as a character predictor to write complete texts. They also have code on github.com/karpathy/char-rnn ready to train / go. You can feed it with a start string and ask for the next characters / words.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.