N-1%DroneMedia Bot

November 14 2020

N-1%DroneMedia is a Twitter bot that posts sentiment data analysed from a news article. It posts daily and only once in the morning. For the moment, the bot is making a single call to The Guardian newspaper API and getting the latest published article at the time of request. The text is then passed onto IBM's Natural Language Processing API to get back sentiment and emotion data based on keywords picked up in the article.

My main idea here is to have some simple automatic means by which I can keep a record of a process and start to look at any trends in the data. I can't really make any inference beyond what these two sources/APIs do; grab the latest news article and chuck it into IBM's machine learning tool. At the time of writing, the sentiment data is looking pretty gloomy and this does seem pertinent when one looks at the articles it is analysing. It is of course in no way representative of news articles in general. Again, the bot is choosing only one source on a daily basis and operating on a single article from that source - usually the headline article from what I can see.

I may well expand on this and set up another bot that uses multiple news sources. There are indeed many approaches that can be taken but at the moment I really don't have any specific objectives either other than record an automatic process. That process being very limited in scope but will enable me to make some observations for future ideas no doubt.

If you are interested in what the bot is tweeting, then you can of course follow it at this Twitter handle.