I, as part of a team, was tasked to create a web accessible experience for IBM, wherein a user could monitor online sentiment surrounding tweets, filtered by location where possible. Front-end designed using HTML & CSS and Back-end designed using NodeRED.
The final web app features a dynamic map with bubbles in various colours indicating its average sentiment score at different locations indicating the origin of the tweets. The radius of the bubbles are proportional to the number tweets and retweets, each indicated with their own set of bubbles. Clicking on a bubble brings up the respective word clouds visualising the maximum of 20 most frequent words tweeted and the most popular locations mentioned. A line graph is also available to show the historical sentiment in the location of interest. A search bar is also available to search for additional handles, hashtags or keywords that may be of interest.
To achieve the final product, I developed my knowledge of HTML, CSS and Javascript, and gained experience working with API’s, namely that of Twitter and IBM’s proprietary machine learning tools. I also developed my ability to work well within a team and took leadership in moments when it was needed to steer the team towards meeting a goal.