Inspiration

The "insider" is somebody with restricted yet highly privileged information. As student-private investors, we realized that a margin can exist only because there is an inherent inequality in information within the stock market because there are both "outsiders" and "insiders". As an outsider, investment is merely testing your luck. It is only when you are an insider when you can start to test your insight, understanding of the market and MINDFULNESS. Mindful gives everybody an equal starting line.

What it does

MINDFUL helps beginner investors obtain useful information more easily to facilitate "thoughtful" investment. It gives users general overview of each company and provides evidence on such statistics. MINDFUL gathers news articles online and tweets in real-time related to companies of interest. It runs sentiment analysis on each text to compute how the world perceives the particular company i.e. the percentage of texts that have positive/neutral/negative view on the company. Then, the articles with strong sentiments are presented so that the user can read the full article and research deeper into the topic.

How we built it

  • Crawler is built with Python and Javascript, and it is running on Google Cloud Platform to gather Tweets in real-time.
  • The server connecting the crawler, database, and the client is written in Express. The server also runs sentiment analysis on the crawled texts using Google Cloud NLP API.
  • Firebase is used as the Database, and the client (front-end) is built using React.

Challenges we ran into

  • Gathering relevant/clean texts
  • Dealing with API limits, crawling issues, etc.
  • Analysing sentiment on texts of diverse quality and length
  • Building intuitive and easy-to-use UX/UI

Accomplishments that we're proud of

  • MINDFUL is an app that fulfills the goal of the project
  • Real-time crawling/sentiment analysis of text data

What we learned

  • Using Google Cloud APIs to develop effective models quickly and easily.
  • Insight on how to make user-friendly service
  • Text-processing skills

What's next for MINDFUL

  • Processing more amount of real-time data yet to be accessible for everybody.
  • Development of more robust sentiment analysis program.
  • Statistical analysis/machine learning to predict the stock price trend based on text data.

(Our team uploaded the API KEY for Firebase/GCP to private Github Repository by mistake. Public repository will be available in the future.)

Share this project:

Updates