ChainCRF for text classification

[aby.s...@gmail.com]

Hello,

   I want to apply the ChainCRF in the Pystruct library to a problem classifying sentences in a resume/CV.

Label fields defined to start with 0 (labels as summary/skills/projects etc).

Please help me how to format the n_features for Unary evidence x.
- n_nodes = sequence number of sentences within a resume
for n_features how should the words in the sentences be represented/formatted?

Thanks

Aby