Sentiment analysis the the algorithmic approach to determining whether text is positive or negative. In other words, when you post something, we can use computers to figure out how positive or negative the language you use is.
You can try it out for any account here .
Just type in any account name and click "Generate".
Here's the sentiment analysis for
New Features
- What feature(s) did you add?
The new feature is the sentiment analysis, displayed in a graph. There was also some code cleanup and integration of build tooling to make the overall code cleaner.
It uses the sentiment npm package to create the scores for each post and comment and then Chart.js to display the results.
The relevant commits:
- Refactor + Webpack/Gulp In this commit, I added Webpack & Gulp to significantly improve the tooling. The code before required me to copy everything into a single html file to get the gitpage working. With gulp and webpack, building is very easy:
gulp buildand everything can be split into files that clearly group the purpose. It also allowed me to get several of the dependencies from npm instead of copy-pasting the source. - Sentiment Analysis This is the commit where I actually added the sentiment analysis with relevant graph.
- Refactor sentiment analysis In this commit, I moved code into separate files to reduce the complexity and scope of the initial file.
Next steps
There are a few directions I could go in, please let me know if any particular feature would be useful.
- Create sentiment analysis search by topic
- Improve the ui: separate pages and localstorage saving of the names (and many more)
- Improve tooling: adding typescript & test framework
Posted on Utopian.io - Rewarding Open Source Contributors