The inspiration behind this project was the power of expert.ai and NLP to gain an understanding of the sentiment of people in the stock market. It is said that one shouldn't invest with emotions, however, we believe that as long as humans are investing, emotions will come into play. Hence, it is important to understand how someone is feeling about a particular stock or the general market. Our project will help investors be more informed about the sentiment of others in the market, and make better decisions for investments.
What it does
Our app gets input such as stock name, investment value, risk level, etc. to generate a stock portfolio. Furthermore, the app uses the Expert.ai API and yahoo-finance API to determine the sentiment of the general market and the specific stock. This will allow users to be more informed.
How we built it
We built it using Django, JS, Python, Yahoo-Finance API, Expert.ai API, jQuery, HTML, CSS, and more. The sentiment value was extracted from Expert.ai's powerful NLP API using the text as input from the Conversation section of Yahoo-Finance. We also got the general news for the market and determined that sentiment as well. The sentiments were then used to create a gauge chart and display Buy or Sell sentiment.
Challenges we ran into
The challenges were finding good freemium financial newsfeed APIs with the quality and quantity of data we need, and the lack of resources. We started out with a team of 5 and reduced it down to 2 in total due to a lack of effort and response from teammates overseas. If we had more time, we would have implemented more features. Furthermore, another challenge we faced was the limitation of expert.ai per request characters. If we exceeded a 10K limit, we got a 403 error. This limited us to make limited requests for news and conversations. Nonetheless, we created more requests to break down one large request.
Accomplishments that we're proud of
We are proud of the fact that we were able to determine and display the sentiment for any stock in the world along with the sentiment of the general news. This allowed us to display our users with more information while investing. Even with all the limitations in data quality, we successfully analyzed sentiments for hundreds of stocks and found a good assessment.
What we learned
Much was gleaned from this hackathon. We gained a good understanding of the expert.ai product and how to leverage its powerful features to build compelling products like this one, and how valuable NLP is. As humans, our primary way of communication is through language. Determining sentiment from language is very powerful and allows powerful applications such as our project to be executed. Furthermore, we learned the use of several popular technologies that helped us create our app.
What's next for Stock-Sentibility
We are going to keep working on our project and implement more features and integrations to allow users to be more informed while investing. If we have the budget, we will invest our money into buying a premium level of Expert.ai API such that we aren't limited in terms of our functionality.