world-triathlon / twitteranalysis Goto Github PK

This is a quick and dirty script to quickly run keyword and sentiment analysis of event tweets via the [Triathlon API] (https://developers.triathlon.org) and MonkeyLearn. It is in no way meant for production usage.

• Step 1: Get a Triathlon API key from the Triathlon API Management Portal
• Step 2: Get a Monkey Learn API key from here
• Step 3: Clone this repo and replace config.php with your API keys obtained in steps 1 & 2 and the event hashtag you wish to analyse (all events may not be available)
• Step 4a: Download composer if not on your local system
• Step 4b: Run composer install
• Step 5: Run the script on the command line with php -f keywords.php to output the results of the keyword analysis

The result of the keyword analysis will produce the top 10 keywords, the number of times they occurred and the relevance (1 being the highest). Note that this analysis was performed on a very small number of tweets for an example.

Keyword | Count | Relevance
weekend | 13 | 0.984
Saturday | 14 | 0.849
American Sarah True | 2 | 0.816
athletes | 9 | 0.751
RACE WEEK | 3 | 0.734
WTS schedule | 3 | 0.734
Alistair Brownlee | 3 | 0.734
WTS podiums | 3 | 0.734
Sweden Stockholm | 3 | 0.734
Jonathan Brownlee | 3 | 0.734

You may also run a sentiment analysis php -f sentiment.php which will yield a result similar to the following:

Positive: 23
Negative: 1
Neutral: 62
Negative Tweet Results
Unfortunately due to illness @Jgomeznoya will not compete in #WTSStockholm this wknd. Wishing him a speedy recovery! https://t.co/Tg2Bl2cnTO