Perform sentiment analysis at the document, sentence, or entity level.
Make sure that you unzip the models in datasets before you try to run the application.
The program will attempt to start a local Stanford server if an existing Stanford server is not running as configured in Config.py.
This program depends upon Stanford CoreNLP. As such it is highly recommended to use 3.9.1 right now. This is a firm requirement when running Agent Annotator on top of it. Also, it is recommended that your instance of CoreNLP be started with 8GB of memory to safely avoid OutOfMemory errors.
The stanfordcorenlp library used was built from the master of https://github.com/Lynten/stanford-corenlp (using setup.py) and not from the commonly available 3.9.1.1 distribution. We need the custom version because it gives us the max_retries which was put into place after the distribution release.
To use the google sentiment, you will have to apply for your google app id. Please see https://developers.google.com/maps/documentation/directions/get-api-key for more information as how to do this.
Both Google and Aylien have limits per day of how many queries can be made. To receive unlimited sentinment, please see the respective websites to consider a paid plan for your particular needs.
Version | About | Date |
---|---|---|
0.1 | Initial Release | 01/17/2019 |