The project is based on stock analysis using both Twitter and AWS API. The program takes your current portfolio and gives recommendations on what to buy and sell based on the data obtained by the Twitter API in tangent with the AWS profile analysis.
Twitter’s API combs through tweets looking for key-value pairs, each word will have a value associated with it between the ranges of -1 and 1. This will be defined in a CSV file. Using these defined values we will develop an ML algorithm. Based on the ratio of negative and positive words, where it lies on the scale between -1 and 1 and combing through multiple tweets, it will develop an average value assigned to that filter. The program will then do an analysis of these stocks, and compare it to your own profile to see if the stock is a fit for you using the AWS API. In addition to recommendations, the program will also display trending stocks based on a live feed of tweets and a breakdown of your current stocks in your portfolio. Finally, the program will also include a portfolio trend that demonstrates your gains and losses from the stocks.
A factor to consider is that Twitter’s API takes data from Twitter’s database itself, reducing the scope of the data that is being gathered. This could easily be overwhelmed due to an influx of tweets so this will need to be regulated. Furthermore, reducing the dataset to purely Twitter logs reduces the scope of the data which could potentially impact the recommendations. In addition, bots from Twitter could potentially influence the program's recommendations as well. When it comes to AWS, latency is another factor to consider for the program and the amount of data available in the AWS database on specific stocks, as some stocks might not be represented in the dataset.
Log in or sign up for Devpost to join the conversation.