SUTDNLP / OpenTargetedSentiment

The implementation of "Neural Networks for Open Domain Targeted Sentiment" based on package https://github.com/SUTDNLP/LibN3L

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

OpenTargetedSentiment

Data and Codes of Neural Networks for Open Domain Targeted Sentiment, EMNLP 2015

To run it, please first compile the codes in the sub-folders of OSentMultiInputFormat and OSentSingleInputFormat.
These codes are implemented based on our neural library LibN3L, thus you should download LibN3L first, modifying the path in CMakeLists.txt.
You will get nine binaries after compiling, and you can follow the commands in script folder to run them.

Please notice the comments in the command files, there are some simple scripts which you should write by yourself, such as split the name entity and sentiment results from one file, or replacing gold-standard name entity results by automatic results in pipeline models.

If you have any question, please email to mason.zms@gmail.com

About

The implementation of "Neural Networks for Open Domain Targeted Sentiment" based on package https://github.com/SUTDNLP/LibN3L


Languages

Language:C++ 98.4%Language:Shell 1.0%Language:CMake 0.6%