Class Projects - Spring 2015


A tool that makes tweeting out pictures with quotes from URLs easy for anyone.

What it does

PicTweet is a web application that is designed to create shareable content from articles across various media sources. Research supports that followers engage with tweets more when images are included. We provide users with a graphic that includes an image from the article with the most relevant quote determined by our natural language processing algorithm. The result will allow individuals the opportunity to include quotations that are longer than 140 characters and even include more hashtags and other users in the tweet itself. This allows for greater contribution to social conversation and better engagement.

How it works

The users visits our site and enters (copy & paste) the article URL. That information is sent to our back-end where the page is scraped and the text is processed by our algorithm. The image and relevant quote are sent back to the presentation layer where they are displayed. At that point, the user has the option of editing the quote before tweeting it out.

Key Technologies (itemized list):


Ruby on Rails


Next Steps

Bookmarklet/Extension: In the future we hope to rework the web application into a bookmarklet or extension to allow for users to simply click on our icon directly from the article and skip the step of copy & pasting the URL

OAuth: Adding authentication so that numerous users can use our site at the same time

Improve NLP: We have worked on finding language cues that help define a powerful quote over the course of the quarter and would like to continue testing and advancing the algorithm.


Monisha Appalaraju, Patricia Gomes, Bhavita Jaiswal, Matt Schley