Daniel Stefanescu

Hearsay

HearSay makes staying on top of your favorite sports teams and players easy with user profiles and fluid question and answering. Our target audience is sports fans across the globe who find it difficult to keep up with a surplus of often conflicting rumors about their favorite teams and players. Our project creates a solution which encapsulates information for the user without neglecting human emotion.

• Github Repository
• Live Demo

UofM Subreddit Automated Analysis

This project analyzes the sentiment of the University of Michigan Subreddit over time and in relation to courses. We compare the information gathered from the subreddit with University polling data to test consistency. Since many students rely on University data when choosing courses, we aim to shed more light on the importance of various metrics as related to student course enjoyment. We found significant results relating University metrics to course sentiment.

• Github Repository
• Paper

Webscraping for Research

I write web scraping programs to automatically collect finance data for researchers at Vanderbilt University and the University of Texas at Austin. For automated authentication and source code parsing, I use various python libraries including Selenium, BeautifulSoup, and Pandas. The programs are designed to be robust and efficient in order to retrieve and process millions of data points.

While I can't post my code publicly to protect authentication details, I have written an example below to scrape details off of my own website!

• Demo Coming Soon!