FDA And Sentiment Analysis

Final Project Proposal

In the modern world, social media is becoming the place to express sentiments. People are using social media as the call for help, a way to report news, a way to shout out things happening around you. It is equally used by people in every walks of life journalists, business leaders, celebrities and also corporate America. I would like to study the FDA recalls, the sentiment expressed on Twitter and its impact on stock prices. For this study, I will be using publicly traded company Whole Foods Market.

Data Collection

• FDA recall report from U.S. Food & Drug Administration. Data is in XML format
• Twitter feeds - Unstructured data
• Stock prices from Yahoo Finance
• Sentiment Lexicon - SentiWordNet

Data Storage

I will be using MongoDB to store the data. Once a collection is created on MongoDB, I will use MapReduce function to create data summary.

Statistical computations

Data will be used for the following analysis

• Recall for given period.
• Words used to express sentiment
• Stock price movement before and after release of recall news

Graphical Reports

• Frequently used words about Whole Foods Market on Twitter.
• Stock price movement before and after release of recall news
• Number of recalls

Challenges expecting

• Downloading data from Twitter and formatting it.
• Data needs to be verified for missing values.
• Using MongoDB and MapReduce for the first time.
• CRUD operations on MongoDB.