The Sentinator

A tool for measuring setiment, well more of a playground trying out different techniques. This analyses @realDonaldTrumps recent tweets and marks them as positive, negative, or neutral.

The Rough Plan

• Pick a subject - Twitter handle - realdonaldtrump - DONE
• Move through time and measure sentiment of tweet. - DONE For Recent Tweets
• Calculate sentiment Value based off historical values with older values decaying in their influence on present value. - NOT DONE
• Record daily changes in sentiment. - NOT DONE

Setup

Python

sudo virtualenv venv
source venv/bin/activate
pip install -r requirements.txt

Java

Download this file http://nlp.stanford.edu/software/stanford-corenlp-full-2016-10-31.zip in this directory nlp/standford.

NLP Examples

Example:

java -cp "*" -Xmx2g edu.stanford.nlp.pipeline.StanfordCoreNLP -annotators tokenize,ssplit,pos,lemma,ner,parse,dcoref -file input.txt

Sentiment Example:

java -cp "*" -Xmx2g edu.stanford.nlp.pipeline.StanfordCoreNLP -annotators tokenize,ssplit,pos,parse,sentiment -file input.txt -outputFormat json

Run

python extractor/extract_tweet.py

Example Output:

Annotation pipeline timing information:
TokenizerAnnotator: 0.1 sec.
WordsToSentencesAnnotator: 0.0 sec.
POSTaggerAnnotator: 0.1 sec.
ParserAnnotator: 3.5 sec.
SentimentAnnotator: 0.2 sec.
TOTAL: 3.9 sec. for 446 tokens at 115.6 tokens/sec.
Pipeline setup: 1.6 sec.
Total time for StanfordCoreNLP pipeline: 5.7 sec.
Positive: 5
Neutral: 8
Negative: 19

Web App

Deployed At: https://thinkjones.github.io/sentinator/#

Sources

• https://medium.freecodecamp.com/serverless-rest-api-with-angular-persistence-and-security-ff274f04e3d0