Click Here to try it out! NOTE: If the site appears slow, it is because you are searching for a stock that has not been previously computed and cached.
For years people have tried to find ways to predict the performance of the stock market in hopes of becoming rich and famous. Many sites on the web offer visitors their recommendations for which stocks one should buy and sell. For this project, over 1 million stock predictions ranging from 2007-2010 were aggragated from around the web and analyzed in order to find out which sites predicted the performance of stocks the most accurately. Use the query to search by stock name, year and site to find out about the quality of predictions of different sites. Developed by Gal Cohen for the undergraduate Database class of Spring 2012, at Rutgers University.
Most impressive ones are bolded
- Paginated Search results (Also improves query time)
- Stock Recommendation CSV Upload
- Sortable Columns for the Search results
- Summary Boxes for Fraction of recommendations that are correct for 1/5/30/365 day intervals - Load with AJAX
- Summary Boxes for Weighted overall Returns - Load with AJAX
- Improved UI for accessiblity and Readability (also removed Prices and Recommenation options)
- About Page
- Source column displays domain name instead of full url
- Added information about database size and Rutgers affiliation
- Domain search field changed to a dynamic dropdown menu updated from the database
- Site Navigation