BiLSTM-CRF Sequence Tagging for E-Commerce Attribute Extraction – Stokastik
In the last post we had used Conditional Random Fields (CRF) to extract attributes from e-commerce product titles and description. CRFs are linear models just like Logistic Regression. The drawback with linear models is that they do not take feature-feature interaction or higher order feature terms into account while building model. Linear models can under-fit on the data while too much non-linearity can lead to over-fitting. Non-linear models such as Neural Networks with proper regularization can be useful.
Instead of Feedforward Neural Networks we can use Recurrent Neural Networks or LSTM networks because they emit an output at each time step i.e. if we are working with text data, then RNN or LSTM outputs a label for each word. So this is how, we can train a sequence tagging model with LSTM.
Read full article from BiLSTM-CRF Sequence Tagging for E-Commerce Attribute Extraction – Stokastik
No comments:
Post a Comment