PRAnalytics

An example of the rating, stock prices, and related news of Google.

Inspiration

With such a versatile economy, it is becoming increasingly difficult for everyday consumers to make important financial decisions. There are a vast number of factors that can go into financial decision making for investors, companies, and most importantly, the people. One factor that is overlooked, both mathematically and theoretically, is public relations.

What it does

Our website analyzes a company's public relations from 30,000 news sources around the world through a Natural Language API to find an arbitrary value of how that company is perceived by the public in the news. This score is then rated on scale that is easy to understand for the consumer. On top of that, the website offers real time information regarding the company and it's financial performance.

How we built it

We used several different APIs that we found for free online to first find the top headlines related to a company being searched for. These article headlines and titles were then ran through the Google Cloud Natural Language API which analyzed the sentiment in the text. Based on the scores received, we compiled this information into a simple score for the company which can be used to interpret whether a company is doing good or bad.

Challenges we ran into

We had trouble finding the correct APIs to use that would fit our needs and the response data would be easy to interpret. We had trouble accessing a few of these APIs due to authentication problems or free tier limitations. We were able to solve a lot of these problems by doing more research on what APIs are available and the documentation of how to use them.

Accomplishments that we're proud of

We managed to create a finished product while implementing many extra features that weren't a part of our MVP.

What we learned

We learned a lot about APIs, how to use them and how to process the data received from them to use for other applications.

What's next for PRAnalytics

In the future, we plan to use our first beta as a small component of a complete database that can predict changes in stock prices for publicly traded companies. We plan to switch the Google Cloud Natural Language API to the AutoML API so that we can train the model ourselves for higher accuracy. We also plan on doing research on better financial APIs that can help us expand in volume and reach.