problem
Direct loading HIT LTP in the speech marking model, for words have word list of words marked the emergence of the mark list of tags is empty phenomenon, the length normally mark list and word list of words the same length, the coexistence marked results.
As shown, an error "Index out of range"
The reason
When the word is empty, LTP speech marking model can not successfully tagged and thrown Index out of range exception.
Solution
Delete all blank lines
- notepad ++ word document opens the document (the program is read from the document and stored in the list) directly delete blank lines
- Preparation of written word filter out blank lines when the result into the document, or filter when reading