The problem Stonks solves

Most conventional stock predictive models are highly dependent on technical analysis of a company's stock and are heavily reliant on crunching numbers and data, often taking into consideration how the stock has performed over time or the history of similar stocks.

This predictive model might be foolproof in a perfect world but in the real world, stock prices and growth/decline rates are often affected by other factors like news and media, which technical analysis often fails to take into account.

Our Solution

To tackle this very predicament, we wanted to build a model that not only takes into consideration the history of the stock, as all generic predictors do but also the current happenings in the world through news and media sources. This is where Stonks steps in, scraping news data from InShorts, a celebrated news source in real-time. Stonks' predictive algorithm identifies the named entities, performs sentiment analysis on the news headlines, and links it to the particular stocks that might be affected by that news using keyword analysis of the news content.

Challenges we ran into

The biggest challenge we ran into was getting accurate and reputable data-sources for finance and cryptocurrency news and prices. We were able to overcome this by scraping data from India's most trusted news source InShorts for finance news, and CryptoControl for crypto news.