Financial Doc Info Visualizer

• 

  User input for sentiment and FLS analysis of any chosen financial report, stock quote, market article etc.

• 

  Pre-prepared snippets of financial reports from several major companies

• 

  Sentiment sentence by sentence analysis with green being positive, gray being neutral, and red being negative

• 

  Pie chart to help easily visualize the general sentiment

• 

  Bar chart to help easily visualize counts of FLS

• 

  FLS sentence by sentence analysis following the same color scheme

Inspiration

This project was inspired by the need to analyze and understand sentiments in financial documents quickly. This can help traders, researchers, civil servants, and citizens to make more informed decisions. This also helps to quantify news articles in a way that can be useful to algorithms.

What it does

This takes text intended to be from financial reports, and does sentence by sentence sentiment and forward looking statement analysis. Sentences are highlighted with color such that positive statements are green, neutral statements are gray, and negative statements are red. Graphs are provided for easy reading, and the project is primarily intended for personal use.

How we built it

The development process involved integrating advanced NLP libraries like spaCy and Transformers. We adopted Streamlit to create an interactive web application interface.

Challenges we ran into

We ran into some challenges with developing good UI and visual interfaces. A number of interactive components required tinkering to work effectively.

Accomplishments that we're proud of

We're proud of producing something that can be a convenient helper to individuals who want to read and quickly grasp financial articles.

What we learned

We learned a lot about good front-end design and visualization. We also learned how to clean and filter data to enable this visualization.

What's next for Financial Doc Info Visualizer

To add functionality to scrape relevant text from the web. This will allow a user to input a link instead of having to copy paste text.

Built With

• matplotlib
• numpy
• pandas
• python
• spacy
• streamlit
• transformers